bs
/
Step_Data_Project_India
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity
You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
10 Commits
1 Branch
0 Tags
145 MiB
 
 
Branch: master
Branches Tags
${ item.name }
Create tag ${ searchTerm }
Create branch ${ searchTerm }
from 'master'
${ noResults }
Step_Data_Project_India/DATA_PREPROCESSING_CODE.ipynb

2882 lines
173 KiB
Raw Permalink Blame History

{
"nbformat": 4,
"nbformat_minor": 0,
"metadata": {
"colab": {
"provenance": [],
"gpuType": "V28"
},
"kernelspec": {
"name": "python3",
"display_name": "Python 3"
},
"language_info": {
"name": "python"
},
"accelerator": "TPU"
},
"cells": [
{
"cell_type": "code",
"source": [
"import os\n",
"\n",
"# Define the path to your folder in Google Drive\n",
"folder_path = '/content/drive/My Drive/Data/iOS/'\n",
"\n",
"# List files in the directory (optional, to verify the files are there)\n",
"print(os.listdir(folder_path))"
],
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/"
},
"id": "l0AqA2w1UaXd",
"outputId": "57108b7e-ec98-483e-f197-2d1795d333d4"
},
"execution_count": null,
"outputs": [
{
"output_type": "stream",
"name": "stdout",
"text": [
"['StepCount01.csv', 'StepCount05.csv', 'StepCount07.csv', 'StepCount09.csv', 'StepCount10.csv', 'StepCount02.csv', 'StepCount08.csv', 'StepCount06.csv', 'StepCount12.csv', 'StepCount03.csv', 'StepCount11.csv', 'StepCount04.csv', 'StepCount20.csv', 'StepCount15.csv', 'StepCount23.csv', 'StepCount17.csv', 'StepCount13.csv', 'StepCount22.csv', 'StepCount24.csv', 'StepCount19.csv', 'StepCount18.csv', 'StepCount14.csv', 'StepCount16.csv', 'StepCount21.csv', 'StepCount29.csv', 'StepCount25.csv', 'StepCount30.csv', 'StepCount26.csv', 'StepCount28.csv', 'StepCount27.csv', 'StepCount31.csv', 'StepCount33.csv', 'StepCount32.csv', 'StepCount34.csv', 'StepCount36.csv', 'StepCount39.csv', 'StepCount38.csv', 'StepCount37.csv', 'StepCount35.csv', 'StepCount42.csv', 'StepCount43.csv', 'StepCount44.csv', 'StepCount41.csv', 'StepCount45.csv', 'StepCount40.csv', 'StepCount46.csv']\n"
]
}
]
},
{
"cell_type": "code",
"source": [
"from google.colab import drive\n",
"drive.mount('/content/drive')"
],
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/"
},
"id": "j-ICoG5Krvsr",
"outputId": "36255c09-8188-464a-92d9-ee581197a96b"
},
"execution_count": null,
"outputs": [
{
"output_type": "stream",
"name": "stdout",
"text": [
"Drive already mounted at /content/drive; to attempt to forcibly remount, call drive.mount(\"/content/drive\", force_remount=True).\n"
]
}
]
},
{
"cell_type": "code",
"source": [
"!pip install openpyxl\n",
"# Install the openpyxl package"
],
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/"
},
"id": "ZprcMokxfBo8",
"outputId": "152a44df-8510-4c2c-e61b-13f1ed35edcb"
},
"execution_count": null,
"outputs": [
{
"output_type": "stream",
"name": "stdout",
"text": [
"Collecting openpyxl\n",
" Downloading openpyxl-3.1.5-py2.py3-none-any.whl.metadata (2.5 kB)\n",
"Collecting et-xmlfile (from openpyxl)\n",
" Downloading et_xmlfile-1.1.0-py3-none-any.whl.metadata (1.8 kB)\n",
"Downloading openpyxl-3.1.5-py2.py3-none-any.whl (250 kB)\n",
"\u001b[2K \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m250.9/250.9 kB\u001b[0m \u001b[31m1.7 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
"\u001b[?25hDownloading et_xmlfile-1.1.0-py3-none-any.whl (4.7 kB)\n",
"Installing collected packages: et-xmlfile, openpyxl\n",
"Successfully installed et-xmlfile-1.1.0 openpyxl-3.1.5\n"
]
}
]
},
{
"cell_type": "markdown",
"source": [
"# FINAL CODE OF 1HR WITHOUT THRESHOLD"
],
"metadata": {
"id": "oY2xhtbJMH4p"
}
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/"
},
"collapsed": true,
"id": "i-kzVCTXtzF4",
"outputId": "4274708b-0e98-416f-9644-ad8de5e809e7"
},
"outputs": [
{
"output_type": "stream",
"name": "stderr",
"text": [
"<ipython-input-5-a4235d7882e2>:15: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['startDate'] = pd.to_datetime(iphone_df['startDate'], format='%Y-%m-%d %H:%M:%S %z')\n",
"<ipython-input-5-a4235d7882e2>:18: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['date'] = iphone_df['startDate'].dt.date\n",
"<ipython-input-5-a4235d7882e2>:19: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['hour'] = iphone_df['startDate'].dt.hour\n"
]
},
{
"output_type": "stream",
"name": "stdout",
"text": [
"/content/drive/My Drive/Data/iOS/StepCount01.csv 0\n"
]
},
{
"output_type": "stream",
"name": "stderr",
"text": [
"<ipython-input-5-a4235d7882e2>:15: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['startDate'] = pd.to_datetime(iphone_df['startDate'], format='%Y-%m-%d %H:%M:%S %z')\n",
"<ipython-input-5-a4235d7882e2>:18: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['date'] = iphone_df['startDate'].dt.date\n",
"<ipython-input-5-a4235d7882e2>:19: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['hour'] = iphone_df['startDate'].dt.hour\n"
]
},
{
"output_type": "stream",
"name": "stdout",
"text": [
"/content/drive/My Drive/Data/iOS/StepCount02.csv 1\n",
"/content/drive/My Drive/Data/iOS/StepCount03.csv 2\n",
"/content/drive/My Drive/Data/iOS/StepCount04.csv 3\n"
]
},
{
"output_type": "stream",
"name": "stderr",
"text": [
"<ipython-input-5-a4235d7882e2>:15: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['startDate'] = pd.to_datetime(iphone_df['startDate'], format='%Y-%m-%d %H:%M:%S %z')\n",
"<ipython-input-5-a4235d7882e2>:18: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['date'] = iphone_df['startDate'].dt.date\n",
"<ipython-input-5-a4235d7882e2>:19: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['hour'] = iphone_df['startDate'].dt.hour\n"
]
},
{
"output_type": "stream",
"name": "stdout",
"text": [
"/content/drive/My Drive/Data/iOS/StepCount05.csv 4\n"
]
},
{
"output_type": "stream",
"name": "stderr",
"text": [
"<ipython-input-5-a4235d7882e2>:15: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['startDate'] = pd.to_datetime(iphone_df['startDate'], format='%Y-%m-%d %H:%M:%S %z')\n",
"<ipython-input-5-a4235d7882e2>:18: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['date'] = iphone_df['startDate'].dt.date\n",
"<ipython-input-5-a4235d7882e2>:19: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['hour'] = iphone_df['startDate'].dt.hour\n"
]
},
{
"output_type": "stream",
"name": "stdout",
"text": [
"/content/drive/My Drive/Data/iOS/StepCount07.csv 5\n"
]
},
{
"output_type": "stream",
"name": "stderr",
"text": [
"<ipython-input-5-a4235d7882e2>:15: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['startDate'] = pd.to_datetime(iphone_df['startDate'], format='%Y-%m-%d %H:%M:%S %z')\n",
"<ipython-input-5-a4235d7882e2>:18: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['date'] = iphone_df['startDate'].dt.date\n",
"<ipython-input-5-a4235d7882e2>:19: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['hour'] = iphone_df['startDate'].dt.hour\n"
]
},
{
"output_type": "stream",
"name": "stdout",
"text": [
"/content/drive/My Drive/Data/iOS/StepCount08.csv 6\n",
"/content/drive/My Drive/Data/iOS/StepCount09.csv 7\n",
"/content/drive/My Drive/Data/iOS/StepCount11.csv 8\n",
"/content/drive/My Drive/Data/iOS/StepCount14.csv 9\n",
"/content/drive/My Drive/Data/iOS/StepCount16.csv 10\n",
"/content/drive/My Drive/Data/iOS/StepCount19.csv 11\n",
"/content/drive/My Drive/Data/iOS/StepCount21.csv 12\n",
"/content/drive/My Drive/Data/iOS/StepCount22.csv 13\n"
]
},
{
"output_type": "stream",
"name": "stderr",
"text": [
"<ipython-input-5-a4235d7882e2>:15: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['startDate'] = pd.to_datetime(iphone_df['startDate'], format='%Y-%m-%d %H:%M:%S %z')\n",
"<ipython-input-5-a4235d7882e2>:18: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['date'] = iphone_df['startDate'].dt.date\n",
"<ipython-input-5-a4235d7882e2>:19: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['hour'] = iphone_df['startDate'].dt.hour\n"
]
},
{
"output_type": "stream",
"name": "stdout",
"text": [
"/content/drive/My Drive/Data/iOS/StepCount23.csv 14\n"
]
},
{
"output_type": "stream",
"name": "stderr",
"text": [
"<ipython-input-5-a4235d7882e2>:15: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['startDate'] = pd.to_datetime(iphone_df['startDate'], format='%Y-%m-%d %H:%M:%S %z')\n",
"<ipython-input-5-a4235d7882e2>:18: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['date'] = iphone_df['startDate'].dt.date\n",
"<ipython-input-5-a4235d7882e2>:19: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['hour'] = iphone_df['startDate'].dt.hour\n"
]
},
{
"output_type": "stream",
"name": "stdout",
"text": [
"/content/drive/My Drive/Data/iOS/StepCount25.csv 15\n",
"/content/drive/My Drive/Data/iOS/StepCount26.csv 16\n"
]
},
{
"output_type": "stream",
"name": "stderr",
"text": [
"<ipython-input-5-a4235d7882e2>:15: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['startDate'] = pd.to_datetime(iphone_df['startDate'], format='%Y-%m-%d %H:%M:%S %z')\n",
"<ipython-input-5-a4235d7882e2>:18: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['date'] = iphone_df['startDate'].dt.date\n",
"<ipython-input-5-a4235d7882e2>:19: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['hour'] = iphone_df['startDate'].dt.hour\n"
]
},
{
"output_type": "stream",
"name": "stdout",
"text": [
"/content/drive/My Drive/Data/iOS/StepCount28.csv 17\n",
"/content/drive/My Drive/Data/iOS/StepCount29.csv 18\n",
"/content/drive/My Drive/Data/iOS/StepCount30.csv 19\n"
]
},
{
"output_type": "stream",
"name": "stderr",
"text": [
"<ipython-input-5-a4235d7882e2>:15: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['startDate'] = pd.to_datetime(iphone_df['startDate'], format='%Y-%m-%d %H:%M:%S %z')\n",
"<ipython-input-5-a4235d7882e2>:18: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['date'] = iphone_df['startDate'].dt.date\n",
"<ipython-input-5-a4235d7882e2>:19: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['hour'] = iphone_df['startDate'].dt.hour\n"
]
},
{
"output_type": "stream",
"name": "stdout",
"text": [
"/content/drive/My Drive/Data/iOS/StepCount33.csv 20\n",
"/content/drive/My Drive/Data/iOS/StepCount34.csv 21\n",
"/content/drive/My Drive/Data/iOS/StepCount35.csv 22\n",
"/content/drive/My Drive/Data/iOS/StepCount36.csv 23\n"
]
},
{
"output_type": "stream",
"name": "stderr",
"text": [
"<ipython-input-5-a4235d7882e2>:15: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['startDate'] = pd.to_datetime(iphone_df['startDate'], format='%Y-%m-%d %H:%M:%S %z')\n",
"<ipython-input-5-a4235d7882e2>:18: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['date'] = iphone_df['startDate'].dt.date\n",
"<ipython-input-5-a4235d7882e2>:19: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['hour'] = iphone_df['startDate'].dt.hour\n"
]
},
{
"output_type": "stream",
"name": "stdout",
"text": [
"/content/drive/My Drive/Data/iOS/StepCount37.csv 24\n"
]
},
{
"output_type": "stream",
"name": "stderr",
"text": [
"<ipython-input-5-a4235d7882e2>:15: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['startDate'] = pd.to_datetime(iphone_df['startDate'], format='%Y-%m-%d %H:%M:%S %z')\n",
"<ipython-input-5-a4235d7882e2>:18: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['date'] = iphone_df['startDate'].dt.date\n",
"<ipython-input-5-a4235d7882e2>:19: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['hour'] = iphone_df['startDate'].dt.hour\n"
]
},
{
"output_type": "stream",
"name": "stdout",
"text": [
"/content/drive/My Drive/Data/iOS/StepCount38.csv 25\n",
"/content/drive/My Drive/Data/iOS/StepCount39.csv 26\n"
]
},
{
"output_type": "stream",
"name": "stderr",
"text": [
"<ipython-input-5-a4235d7882e2>:15: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['startDate'] = pd.to_datetime(iphone_df['startDate'], format='%Y-%m-%d %H:%M:%S %z')\n",
"<ipython-input-5-a4235d7882e2>:18: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['date'] = iphone_df['startDate'].dt.date\n",
"<ipython-input-5-a4235d7882e2>:19: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['hour'] = iphone_df['startDate'].dt.hour\n"
]
},
{
"output_type": "stream",
"name": "stdout",
"text": [
"/content/drive/My Drive/Data/iOS/StepCount40.csv 27\n"
]
},
{
"output_type": "stream",
"name": "stderr",
"text": [
"<ipython-input-5-a4235d7882e2>:15: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['startDate'] = pd.to_datetime(iphone_df['startDate'], format='%Y-%m-%d %H:%M:%S %z')\n",
"<ipython-input-5-a4235d7882e2>:18: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['date'] = iphone_df['startDate'].dt.date\n",
"<ipython-input-5-a4235d7882e2>:19: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['hour'] = iphone_df['startDate'].dt.hour\n"
]
},
{
"output_type": "stream",
"name": "stdout",
"text": [
"/content/drive/My Drive/Data/iOS/StepCount41.csv 28\n",
"/content/drive/My Drive/Data/iOS/StepCount43.csv 29\n",
"/content/drive/My Drive/Data/iOS/StepCount44.csv 30\n",
"/content/drive/My Drive/Data/iOS/StepCount45.csv 31\n",
" date Hour_0 Hour_1 Hour_2 Hour_3 Hour_4 Hour_5 Hour_6 \\\n",
"0 2017-07-20 False False False False False False False \n",
"1 2017-07-21 False False False False False False False \n",
"2 2017-07-22 True True False False False False False \n",
"3 2017-07-23 False False False False False False False \n",
"4 2017-07-24 True True False False False False False \n",
"... ... ... ... ... ... ... ... ... \n",
"36480 2020-06-09 True False False False False False False \n",
"36481 2020-06-10 False False True False False False False \n",
"36482 2020-06-11 True False False False False False True \n",
"36483 2020-06-12 True False False False False False False \n",
"36484 2020-06-13 True False False False False False False \n",
"\n",
" Hour_7 Hour_8 ... Month Year DayOfWeek_Friday DayOfWeek_Monday \\\n",
"0 False False ... 7 2017 False False \n",
"1 True True ... 7 2017 True False \n",
"2 True True ... 7 2017 False False \n",
"3 False True ... 7 2017 False False \n",
"4 False True ... 7 2017 False True \n",
"... ... ... ... ... ... ... ... \n",
"36480 False False ... 6 2020 False False \n",
"36481 False False ... 6 2020 False False \n",
"36482 True True ... 6 2020 False False \n",
"36483 False False ... 6 2020 True False \n",
"36484 False False ... 6 2020 False False \n",
"\n",
" DayOfWeek_Saturday DayOfWeek_Sunday DayOfWeek_Thursday \\\n",
"0 False False True \n",
"1 False False False \n",
"2 True False False \n",
"3 False True False \n",
"4 False False False \n",
"... ... ... ... \n",
"36480 False False False \n",
"36481 False False False \n",
"36482 False False True \n",
"36483 False False False \n",
"36484 True False False \n",
"\n",
" DayOfWeek_Tuesday DayOfWeek_Wednesday user \n",
"0 False False 0 \n",
"1 False False 0 \n",
"2 False False 0 \n",
"3 False False 0 \n",
"4 False False 0 \n",
"... ... ... ... \n",
"36480 True False 31 \n",
"36481 False True 31 \n",
"36482 False False 31 \n",
"36483 False False 31 \n",
"36484 False False 31 \n",
"\n",
"[36485 rows x 35 columns]\n"
]
}
],
"source": [
"import pandas as pd\n",
"\n",
"def process_file(file_path, user_label):\n",
"\n",
" # Load the dataset\n",
" df = pd.read_csv(file_path, delimiter=';')\n",
"\n",
" # Step 1: Filter for iPhone devices\n",
" iphone_df = df[df['device'].str.contains('iPhone', na=False)] # Treat NaN as False\n",
"\n",
" # Step 2: Select the desired columns\n",
" result = iphone_df[['startDate', 'endDate', 'value']]\n",
"\n",
" # Step 3: Convert startDate to datetime\n",
" iphone_df['startDate'] = pd.to_datetime(iphone_df['startDate'], format='%Y-%m-%d %H:%M:%S %z')\n",
"\n",
" # Step 4: Extract date and hour\n",
" iphone_df['date'] = iphone_df['startDate'].dt.date\n",
" iphone_df['hour'] = iphone_df['startDate'].dt.hour\n",
"\n",
" # Step 5: Group by date and hour, then sum the values\n",
" hourly_sum = iphone_df.groupby(['date', 'hour'])['value'].sum().reset_index()\n",
"\n",
" # Step 6: Pivot the data to get one row per day with 24 columns for each hour\n",
" pivot_table = hourly_sum.pivot(index='date', columns='hour', values='value').fillna(0)\n",
"\n",
" # Step 7: Rename columns to reflect hours\n",
" pivot_table.columns = [f'Hour_{i}' for i in pivot_table.columns]\n",
"\n",
" # Step 8: Reset index to have 'date' as a column instead of index\n",
" pivot_table.reset_index(inplace=True)\n",
"\n",
" # Step 9: Add day of the week, month, and year columns\n",
" pivot_table['DayOfWeek'] = pd.to_datetime(pivot_table['date']).dt.day_name()\n",
" pivot_table['Month'] = pd.to_datetime(pivot_table['date']).dt.month\n",
" pivot_table['Year'] = pd.to_datetime(pivot_table['date']).dt.year\n",
"\n",
" # Step 10: One-hot encode the 'DayOfWeek' column\n",
" pivot_table = pd.concat([pivot_table, pd.get_dummies(pivot_table['DayOfWeek'], prefix='DayOfWeek')], axis=1)\n",
"\n",
" # Step 11: Convert hourly values to binary (True if > 0, else False)\n",
" for col in pivot_table.columns[1:25]: # Skip the 'date' column and focus on hours\n",
" pivot_table[col] = pivot_table[col].apply(lambda x: True if x > 0 else False)\n",
"\n",
" # Step 12: Add 'user' column with the specified user label\n",
" pivot_table['user'] = user_label\n",
" # Print which file is currently being processed\n",
" print(file_path,user_label)\n",
" # Step 13: Drop the 'DayOfWeek' column\n",
" pivot_table.drop(columns=['DayOfWeek'], inplace=True)\n",
"\n",
" return pivot_table\n",
"\n",
"# List of files to skip\n",
"files_to_skip = {'StepCount06.csv','StepCount10.csv','StepCount12.csv', 'StepCount13.csv', 'StepCount15.csv', 'StepCount17.csv',\n",
" 'StepCount18.csv', 'StepCount20.csv', 'StepCount24.csv','StepCount27.csv', 'StepCount31.csv','StepCount32.csv',\n",
" 'StepCount42.csv', 'StepCount46.csv'}\n",
"\n",
"# Generate file paths, skipping specified files\n",
"file_paths = [f'/content/drive/My Drive/Data/iOS/StepCount{i:02d}.csv' for i in range(1, 47)\n",
" if f'StepCount{i:02d}.csv' not in files_to_skip]\n",
"\n",
"# Generate user labels based on file index\n",
"user_labels = list(range(len(file_paths)))\n",
"\n",
"\n",
"# Process each file with its corresponding user label and concatenate the results\n",
"processed_dfs = [process_file(file_path, user_label) for file_path, user_label in zip(file_paths, user_labels)]\n",
"combined_df = pd.concat(processed_dfs, ignore_index=True)\n",
"\n",
"# Save the combined DataFrame to a new Excel file\n",
"updated_file_path = '/content/combined_aggregated_data.xlsx'\n",
"combined_df.to_excel(updated_file_path, index=False)\n",
"\n",
"# Print the final DataFrame\n",
"print(combined_df)\n"
]
},
{
"cell_type": "markdown",
"source": [
"\n",
"# 15MIN WITHOUT THRESHOLD"
],
"metadata": {
"id": "yGsfgd6gL2zD"
}
},
{
"cell_type": "code",
"source": [
"import pandas as pd\n",
"import numpy as np\n",
"\n",
"def process_file(file_path, user_label):\n",
" # Load the dataset\n",
" df = pd.read_csv(file_path, delimiter=';')\n",
"\n",
" # Filter for iPhone devices\n",
" iphone_df = df[df['device'].str.contains('iPhone', na=False)]\n",
"\n",
" # Convert startDate to datetime\n",
" iphone_df['startDate'] = pd.to_datetime(iphone_df['startDate'], format='%Y-%m-%d %H:%M:%S %z')\n",
"\n",
" # Round down the startDate to the nearest 15-minute interval\n",
" iphone_df['15min_interval'] = iphone_df['startDate'].dt.floor('15T')\n",
"\n",
" # Extract date, time, year, and month for 15-minute intervals\n",
" iphone_df['date'] = iphone_df['15min_interval'].dt.date\n",
" iphone_df['time'] = iphone_df['15min_interval'].dt.time\n",
" iphone_df['Year'] = iphone_df['15min_interval'].dt.year\n",
" iphone_df['Month'] = iphone_df['15min_interval'].dt.month\n",
"\n",
" # Group by date, time, year, and month, then sum the values\n",
"\n",
"\n",
" interval_sum = iphone_df.groupby(['date', 'time', 'Year', 'Month'])['value'].sum().reset_index()\n",
"\n",
" # Create a full range of 15-minute intervals (00:00:00 to 23:45:00)\n",
" full_time_range = pd.date_range('00:00', '23:45', freq='15T').time\n",
"\n",
" # Pivot the data to get one row per day with columns for each 15-minute interval\n",
" pivot_table = interval_sum.pivot(index=['date', 'Year', 'Month'], columns='time', values='value').fillna(0)\n",
"\n",
" # Reindex to include all possible 15-minute intervals\n",
" pivot_table = pivot_table.reindex(columns=full_time_range, fill_value=0)\n",
"\n",
" # Rename columns to reflect 15-minute intervals\n",
" pivot_table.columns = [f'{str(col)}' for col in pivot_table.columns]\n",
"\n",
" # Convert interval values to boolean (True if > 0, else False)\n",
" pivot_table = pivot_table.apply(lambda col: col != 0, axis=0)\n",
"\n",
" # Reset index to have 'date', 'Year', and 'Month' as columns instead of index\n",
" pivot_table.reset_index(inplace=True)\n",
"\n",
" # Add day of the week\n",
" pivot_table['DayOfWeek'] = pd.to_datetime(pivot_table['date']).dt.day_name()\n",
"\n",
" # One-hot encode the 'DayOfWeek' column\n",
" pivot_table = pd.concat([pivot_table, pd.get_dummies(pivot_table['DayOfWeek'], prefix='DayOfWeek')], axis=1)\n",
"\n",
" # Add a user column with the specified user label\n",
" pivot_table['user'] = user_label\n",
"\n",
" # Print which file is currently being processed\n",
" print(f\"Processing file: {file_path}, User label: {user_label}\")\n",
"\n",
" return pivot_table\n",
"\n",
"# List of files to skip\n",
"files_to_skip = {'StepCount06.csv','StepCount10.csv','StepCount12.csv', 'StepCount13.csv', 'StepCount15.csv', 'StepCount17.csv',\n",
" 'StepCount18.csv', 'StepCount20.csv', 'StepCount24.csv', 'StepCount27.csv','StepCount31.csv','StepCount32.csv',\n",
" 'StepCount42.csv', 'StepCount46.csv'}\n",
"\n",
"# Generate file paths, skipping specified files\n",
"file_paths = [f'/content/drive/My Drive/Data/iOS/StepCount{i:02d}.csv' for i in range(1, 47)\n",
" if f'StepCount{i:02d}.csv' not in files_to_skip]\n",
"\n",
"# Generate user labels based on file index\n",
"user_labels = list(range(len(file_paths)))\n",
"\n",
"# Process each file with its corresponding user label and concatenate the results\n",
"processed_dfs = [process_file(file_path, user_label) for file_path, user_label in zip(file_paths, user_labels)]\n",
"combined_df = pd.concat(processed_dfs, ignore_index=True)\n",
"\n",
"# Save the combined DataFrame to a new Excel file\n",
"updated_file_path = '/content/combined_aggregated_data_15min_without_threshold.xlsx'\n",
"combined_df.to_excel(updated_file_path, index=False)\n",
"\n",
"# Print the final DataFrame\n",
"print(combined_df)\n"
],
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/"
},
"id": "0wtxqkYAhLMe",
"outputId": "43e7e468-8cc3-4864-edaa-1b11df3349f8",
"collapsed": true
},
"execution_count": null,
"outputs": [
{
"output_type": "stream",
"name": "stderr",
"text": [
"<ipython-input-4-36a1c351e706>:12: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['startDate'] = pd.to_datetime(iphone_df['startDate'], format='%Y-%m-%d %H:%M:%S %z')\n",
"<ipython-input-4-36a1c351e706>:15: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['15min_interval'] = iphone_df['startDate'].dt.floor('15T')\n",
"<ipython-input-4-36a1c351e706>:18: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['date'] = iphone_df['15min_interval'].dt.date\n",
"<ipython-input-4-36a1c351e706>:19: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['time'] = iphone_df['15min_interval'].dt.time\n",
"<ipython-input-4-36a1c351e706>:20: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['Year'] = iphone_df['15min_interval'].dt.year\n",
"<ipython-input-4-36a1c351e706>:21: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['Month'] = iphone_df['15min_interval'].dt.month\n"
]
},
{
"output_type": "stream",
"name": "stdout",
"text": [
"Processing file: /content/drive/My Drive/Data/iOS/StepCount01.csv, User label: 0\n"
]
},
{
"output_type": "stream",
"name": "stderr",
"text": [
"<ipython-input-4-36a1c351e706>:12: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['startDate'] = pd.to_datetime(iphone_df['startDate'], format='%Y-%m-%d %H:%M:%S %z')\n",
"<ipython-input-4-36a1c351e706>:15: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['15min_interval'] = iphone_df['startDate'].dt.floor('15T')\n",
"<ipython-input-4-36a1c351e706>:18: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['date'] = iphone_df['15min_interval'].dt.date\n",
"<ipython-input-4-36a1c351e706>:19: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['time'] = iphone_df['15min_interval'].dt.time\n",
"<ipython-input-4-36a1c351e706>:20: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['Year'] = iphone_df['15min_interval'].dt.year\n",
"<ipython-input-4-36a1c351e706>:21: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['Month'] = iphone_df['15min_interval'].dt.month\n"
]
},
{
"output_type": "stream",
"name": "stdout",
"text": [
"Processing file: /content/drive/My Drive/Data/iOS/StepCount02.csv, User label: 1\n",
"Processing file: /content/drive/My Drive/Data/iOS/StepCount03.csv, User label: 2\n",
"Processing file: /content/drive/My Drive/Data/iOS/StepCount04.csv, User label: 3\n"
]
},
{
"output_type": "stream",
"name": "stderr",
"text": [
"<ipython-input-4-36a1c351e706>:12: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['startDate'] = pd.to_datetime(iphone_df['startDate'], format='%Y-%m-%d %H:%M:%S %z')\n",
"<ipython-input-4-36a1c351e706>:15: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['15min_interval'] = iphone_df['startDate'].dt.floor('15T')\n",
"<ipython-input-4-36a1c351e706>:18: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['date'] = iphone_df['15min_interval'].dt.date\n",
"<ipython-input-4-36a1c351e706>:19: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['time'] = iphone_df['15min_interval'].dt.time\n",
"<ipython-input-4-36a1c351e706>:20: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['Year'] = iphone_df['15min_interval'].dt.year\n",
"<ipython-input-4-36a1c351e706>:21: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['Month'] = iphone_df['15min_interval'].dt.month\n"
]
},
{
"output_type": "stream",
"name": "stdout",
"text": [
"Processing file: /content/drive/My Drive/Data/iOS/StepCount05.csv, User label: 4\n"
]
},
{
"output_type": "stream",
"name": "stderr",
"text": [
"<ipython-input-4-36a1c351e706>:12: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['startDate'] = pd.to_datetime(iphone_df['startDate'], format='%Y-%m-%d %H:%M:%S %z')\n",
"<ipython-input-4-36a1c351e706>:15: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['15min_interval'] = iphone_df['startDate'].dt.floor('15T')\n",
"<ipython-input-4-36a1c351e706>:18: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['date'] = iphone_df['15min_interval'].dt.date\n",
"<ipython-input-4-36a1c351e706>:19: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['time'] = iphone_df['15min_interval'].dt.time\n",
"<ipython-input-4-36a1c351e706>:20: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['Year'] = iphone_df['15min_interval'].dt.year\n",
"<ipython-input-4-36a1c351e706>:21: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['Month'] = iphone_df['15min_interval'].dt.month\n"
]
},
{
"output_type": "stream",
"name": "stdout",
"text": [
"Processing file: /content/drive/My Drive/Data/iOS/StepCount07.csv, User label: 5\n"
]
},
{
"output_type": "stream",
"name": "stderr",
"text": [
"<ipython-input-4-36a1c351e706>:12: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['startDate'] = pd.to_datetime(iphone_df['startDate'], format='%Y-%m-%d %H:%M:%S %z')\n",
"<ipython-input-4-36a1c351e706>:15: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['15min_interval'] = iphone_df['startDate'].dt.floor('15T')\n",
"<ipython-input-4-36a1c351e706>:18: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['date'] = iphone_df['15min_interval'].dt.date\n",
"<ipython-input-4-36a1c351e706>:19: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['time'] = iphone_df['15min_interval'].dt.time\n",
"<ipython-input-4-36a1c351e706>:20: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['Year'] = iphone_df['15min_interval'].dt.year\n",
"<ipython-input-4-36a1c351e706>:21: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['Month'] = iphone_df['15min_interval'].dt.month\n"
]
},
{
"output_type": "stream",
"name": "stdout",
"text": [
"Processing file: /content/drive/My Drive/Data/iOS/StepCount08.csv, User label: 6\n",
"Processing file: /content/drive/My Drive/Data/iOS/StepCount09.csv, User label: 7\n",
"Processing file: /content/drive/My Drive/Data/iOS/StepCount11.csv, User label: 8\n",
"Processing file: /content/drive/My Drive/Data/iOS/StepCount14.csv, User label: 9\n",
"Processing file: /content/drive/My Drive/Data/iOS/StepCount16.csv, User label: 10\n",
"Processing file: /content/drive/My Drive/Data/iOS/StepCount19.csv, User label: 11\n",
"Processing file: /content/drive/My Drive/Data/iOS/StepCount21.csv, User label: 12\n",
"Processing file: /content/drive/My Drive/Data/iOS/StepCount22.csv, User label: 13\n"
]
},
{
"output_type": "stream",
"name": "stderr",
"text": [
"<ipython-input-4-36a1c351e706>:12: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['startDate'] = pd.to_datetime(iphone_df['startDate'], format='%Y-%m-%d %H:%M:%S %z')\n",
"<ipython-input-4-36a1c351e706>:15: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['15min_interval'] = iphone_df['startDate'].dt.floor('15T')\n",
"<ipython-input-4-36a1c351e706>:18: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['date'] = iphone_df['15min_interval'].dt.date\n",
"<ipython-input-4-36a1c351e706>:19: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['time'] = iphone_df['15min_interval'].dt.time\n",
"<ipython-input-4-36a1c351e706>:20: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['Year'] = iphone_df['15min_interval'].dt.year\n",
"<ipython-input-4-36a1c351e706>:21: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['Month'] = iphone_df['15min_interval'].dt.month\n"
]
},
{
"output_type": "stream",
"name": "stdout",
"text": [
"Processing file: /content/drive/My Drive/Data/iOS/StepCount23.csv, User label: 14\n"
]
},
{
"output_type": "stream",
"name": "stderr",
"text": [
"<ipython-input-4-36a1c351e706>:12: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['startDate'] = pd.to_datetime(iphone_df['startDate'], format='%Y-%m-%d %H:%M:%S %z')\n",
"<ipython-input-4-36a1c351e706>:15: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['15min_interval'] = iphone_df['startDate'].dt.floor('15T')\n",
"<ipython-input-4-36a1c351e706>:18: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['date'] = iphone_df['15min_interval'].dt.date\n",
"<ipython-input-4-36a1c351e706>:19: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['time'] = iphone_df['15min_interval'].dt.time\n",
"<ipython-input-4-36a1c351e706>:20: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['Year'] = iphone_df['15min_interval'].dt.year\n",
"<ipython-input-4-36a1c351e706>:21: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['Month'] = iphone_df['15min_interval'].dt.month\n"
]
},
{
"output_type": "stream",
"name": "stdout",
"text": [
"Processing file: /content/drive/My Drive/Data/iOS/StepCount25.csv, User label: 15\n",
"Processing file: /content/drive/My Drive/Data/iOS/StepCount26.csv, User label: 16\n"
]
},
{
"output_type": "stream",
"name": "stderr",
"text": [
"<ipython-input-4-36a1c351e706>:12: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['startDate'] = pd.to_datetime(iphone_df['startDate'], format='%Y-%m-%d %H:%M:%S %z')\n",
"<ipython-input-4-36a1c351e706>:15: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['15min_interval'] = iphone_df['startDate'].dt.floor('15T')\n",
"<ipython-input-4-36a1c351e706>:18: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['date'] = iphone_df['15min_interval'].dt.date\n",
"<ipython-input-4-36a1c351e706>:19: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['time'] = iphone_df['15min_interval'].dt.time\n",
"<ipython-input-4-36a1c351e706>:20: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['Year'] = iphone_df['15min_interval'].dt.year\n",
"<ipython-input-4-36a1c351e706>:21: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['Month'] = iphone_df['15min_interval'].dt.month\n"
]
},
{
"output_type": "stream",
"name": "stdout",
"text": [
"Processing file: /content/drive/My Drive/Data/iOS/StepCount28.csv, User label: 17\n",
"Processing file: /content/drive/My Drive/Data/iOS/StepCount29.csv, User label: 18\n",
"Processing file: /content/drive/My Drive/Data/iOS/StepCount30.csv, User label: 19\n"
]
},
{
"output_type": "stream",
"name": "stderr",
"text": [
"<ipython-input-4-36a1c351e706>:12: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['startDate'] = pd.to_datetime(iphone_df['startDate'], format='%Y-%m-%d %H:%M:%S %z')\n",
"<ipython-input-4-36a1c351e706>:15: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['15min_interval'] = iphone_df['startDate'].dt.floor('15T')\n",
"<ipython-input-4-36a1c351e706>:18: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['date'] = iphone_df['15min_interval'].dt.date\n",
"<ipython-input-4-36a1c351e706>:19: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['time'] = iphone_df['15min_interval'].dt.time\n",
"<ipython-input-4-36a1c351e706>:20: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['Year'] = iphone_df['15min_interval'].dt.year\n",
"<ipython-input-4-36a1c351e706>:21: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['Month'] = iphone_df['15min_interval'].dt.month\n"
]
},
{
"output_type": "stream",
"name": "stdout",
"text": [
"Processing file: /content/drive/My Drive/Data/iOS/StepCount33.csv, User label: 20\n",
"Processing file: /content/drive/My Drive/Data/iOS/StepCount34.csv, User label: 21\n",
"Processing file: /content/drive/My Drive/Data/iOS/StepCount35.csv, User label: 22\n",
"Processing file: /content/drive/My Drive/Data/iOS/StepCount36.csv, User label: 23\n"
]
},
{
"output_type": "stream",
"name": "stderr",
"text": [
"<ipython-input-4-36a1c351e706>:12: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['startDate'] = pd.to_datetime(iphone_df['startDate'], format='%Y-%m-%d %H:%M:%S %z')\n",
"<ipython-input-4-36a1c351e706>:15: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['15min_interval'] = iphone_df['startDate'].dt.floor('15T')\n",
"<ipython-input-4-36a1c351e706>:18: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['date'] = iphone_df['15min_interval'].dt.date\n",
"<ipython-input-4-36a1c351e706>:19: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['time'] = iphone_df['15min_interval'].dt.time\n",
"<ipython-input-4-36a1c351e706>:20: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['Year'] = iphone_df['15min_interval'].dt.year\n",
"<ipython-input-4-36a1c351e706>:21: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['Month'] = iphone_df['15min_interval'].dt.month\n"
]
},
{
"output_type": "stream",
"name": "stdout",
"text": [
"Processing file: /content/drive/My Drive/Data/iOS/StepCount37.csv, User label: 24\n"
]
},
{
"output_type": "stream",
"name": "stderr",
"text": [
"<ipython-input-4-36a1c351e706>:12: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['startDate'] = pd.to_datetime(iphone_df['startDate'], format='%Y-%m-%d %H:%M:%S %z')\n",
"<ipython-input-4-36a1c351e706>:15: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['15min_interval'] = iphone_df['startDate'].dt.floor('15T')\n",
"<ipython-input-4-36a1c351e706>:18: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['date'] = iphone_df['15min_interval'].dt.date\n",
"<ipython-input-4-36a1c351e706>:19: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['time'] = iphone_df['15min_interval'].dt.time\n",
"<ipython-input-4-36a1c351e706>:20: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['Year'] = iphone_df['15min_interval'].dt.year\n",
"<ipython-input-4-36a1c351e706>:21: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['Month'] = iphone_df['15min_interval'].dt.month\n"
]
},
{
"output_type": "stream",
"name": "stdout",
"text": [
"Processing file: /content/drive/My Drive/Data/iOS/StepCount38.csv, User label: 25\n",
"Processing file: /content/drive/My Drive/Data/iOS/StepCount39.csv, User label: 26\n"
]
},
{
"output_type": "stream",
"name": "stderr",
"text": [
"<ipython-input-4-36a1c351e706>:12: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['startDate'] = pd.to_datetime(iphone_df['startDate'], format='%Y-%m-%d %H:%M:%S %z')\n",
"<ipython-input-4-36a1c351e706>:15: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['15min_interval'] = iphone_df['startDate'].dt.floor('15T')\n",
"<ipython-input-4-36a1c351e706>:18: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['date'] = iphone_df['15min_interval'].dt.date\n",
"<ipython-input-4-36a1c351e706>:19: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['time'] = iphone_df['15min_interval'].dt.time\n",
"<ipython-input-4-36a1c351e706>:20: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['Year'] = iphone_df['15min_interval'].dt.year\n",
"<ipython-input-4-36a1c351e706>:21: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['Month'] = iphone_df['15min_interval'].dt.month\n"
]
},
{
"output_type": "stream",
"name": "stdout",
"text": [
"Processing file: /content/drive/My Drive/Data/iOS/StepCount40.csv, User label: 27\n"
]
},
{
"output_type": "stream",
"name": "stderr",
"text": [
"<ipython-input-4-36a1c351e706>:12: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['startDate'] = pd.to_datetime(iphone_df['startDate'], format='%Y-%m-%d %H:%M:%S %z')\n",
"<ipython-input-4-36a1c351e706>:15: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['15min_interval'] = iphone_df['startDate'].dt.floor('15T')\n",
"<ipython-input-4-36a1c351e706>:18: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['date'] = iphone_df['15min_interval'].dt.date\n",
"<ipython-input-4-36a1c351e706>:19: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['time'] = iphone_df['15min_interval'].dt.time\n",
"<ipython-input-4-36a1c351e706>:20: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['Year'] = iphone_df['15min_interval'].dt.year\n",
"<ipython-input-4-36a1c351e706>:21: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['Month'] = iphone_df['15min_interval'].dt.month\n"
]
},
{
"output_type": "stream",
"name": "stdout",
"text": [
"Processing file: /content/drive/My Drive/Data/iOS/StepCount41.csv, User label: 28\n",
"Processing file: /content/drive/My Drive/Data/iOS/StepCount43.csv, User label: 29\n",
"Processing file: /content/drive/My Drive/Data/iOS/StepCount44.csv, User label: 30\n",
"Processing file: /content/drive/My Drive/Data/iOS/StepCount45.csv, User label: 31\n",
" date Year Month 00:00:00 00:15:00 00:30:00 00:45:00 \\\n",
"0 2017-07-20 2017 7 False False False False \n",
"1 2017-07-21 2017 7 False False False False \n",
"2 2017-07-22 2017 7 False False False True \n",
"3 2017-07-23 2017 7 False False False False \n",
"4 2017-07-24 2017 7 False True False False \n",
"... ... ... ... ... ... ... ... \n",
"36480 2020-06-09 2020 6 False False True False \n",
"36481 2020-06-10 2020 6 False False False False \n",
"36482 2020-06-11 2020 6 False True False False \n",
"36483 2020-06-12 2020 6 False False True False \n",
"36484 2020-06-13 2020 6 False False True False \n",
"\n",
" 01:00:00 01:15:00 01:30:00 ... 23:45:00 DayOfWeek \\\n",
"0 False False False ... False Thursday \n",
"1 False False False ... False Friday \n",
"2 True False False ... False Saturday \n",
"3 False False False ... False Sunday \n",
"4 False False True ... False Monday \n",
"... ... ... ... ... ... ... \n",
"36480 False False False ... False Tuesday \n",
"36481 False False False ... False Wednesday \n",
"36482 False False False ... False Thursday \n",
"36483 False False False ... False Friday \n",
"36484 False False False ... False Saturday \n",
"\n",
" DayOfWeek_Friday DayOfWeek_Monday DayOfWeek_Saturday \\\n",
"0 False False False \n",
"1 True False False \n",
"2 False False True \n",
"3 False False False \n",
"4 False True False \n",
"... ... ... ... \n",
"36480 False False False \n",
"36481 False False False \n",
"36482 False False False \n",
"36483 True False False \n",
"36484 False False True \n",
"\n",
" DayOfWeek_Sunday DayOfWeek_Thursday DayOfWeek_Tuesday \\\n",
"0 False True False \n",
"1 False False False \n",
"2 False False False \n",
"3 True False False \n",
"4 False False False \n",
"... ... ... ... \n",
"36480 False False True \n",
"36481 False False False \n",
"36482 False True False \n",
"36483 False False False \n",
"36484 False False False \n",
"\n",
" DayOfWeek_Wednesday user \n",
"0 False 0 \n",
"1 False 0 \n",
"2 False 0 \n",
"3 False 0 \n",
"4 False 0 \n",
"... ... ... \n",
"36480 False 31 \n",
"36481 True 31 \n",
"36482 False 31 \n",
"36483 False 31 \n",
"36484 False 31 \n",
"\n",
"[36485 rows x 108 columns]\n"
]
}
]
},
{
"cell_type": "code",
"source": [
"user_counts = combined_df['user'].value_counts()\n",
"\n",
"# Display the count of each user\n",
"print(user_counts.sort_index())"
],
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/"
},
"id": "xmiiOmq_G0rJ",
"outputId": "e783652e-168b-487d-a272-74ca465a8b69"
},
"execution_count": null,
"outputs": [
{
"output_type": "stream",
"name": "stdout",
"text": [
"user\n",
"0 1025\n",
"1 1713\n",
"2 796\n",
"3 889\n",
"4 1656\n",
"5 498\n",
"6 880\n",
"7 1094\n",
"8 954\n",
"9 1657\n",
"10 1584\n",
"11 1561\n",
"12 1513\n",
"13 802\n",
"14 1388\n",
"15 1058\n",
"16 782\n",
"17 1155\n",
"18 810\n",
"19 1112\n",
"20 1555\n",
"21 1362\n",
"22 656\n",
"23 1289\n",
"24 829\n",
"25 1623\n",
"26 568\n",
"27 1621\n",
"28 1154\n",
"29 664\n",
"30 976\n",
"31 1261\n",
"Name: count, dtype: int64\n"
]
}
]
},
{
"cell_type": "markdown",
"source": [
"**FINAL CODE OF 15MIN WITH THRESHOLD**"
],
"metadata": {
"id": "iJ4j1YoZ7KbJ"
}
},
{
"cell_type": "code",
"source": [
"import pandas as pd\n",
"\n",
"def process_file(file_path, user_label):\n",
" # Load the dataset\n",
" df = pd.read_csv(file_path, delimiter=';')\n",
"\n",
" # Step 1: Filter for iPhone devices\n",
" iphone_df = df[df['device'].str.contains('iPhone', na=False)] # Treat NaN as False\n",
"\n",
" # Step 2: Select the desired columns\n",
" result = iphone_df[['startDate', 'endDate', 'value']]\n",
"\n",
" # Step 3: Convert startDate to datetime\n",
" iphone_df['startDate'] = pd.to_datetime(iphone_df['startDate'], format='%Y-%m-%d %H:%M:%S %z')\n",
"\n",
" # Step 4: Round down the startDate to the nearest 15-minute interval\n",
" iphone_df['15min_interval'] = iphone_df['startDate'].dt.floor('15T')\n",
"\n",
" # Step 5: Extract date and time\n",
" iphone_df['date'] = iphone_df['15min_interval'].dt.date\n",
" iphone_df['time'] = iphone_df['15min_interval'].dt.time\n",
"\n",
" # Step 6: Group by date and time, then sum the values for 15-minute intervals\n",
" iphone_df_filtered = iphone_df[iphone_df['value'] > 25].dropna(subset=['value'])\n",
" interval_sum = iphone_df.groupby(['date', 'time'])['value'].sum().reset_index()\n",
"\n",
" # Step 7: Pivot the data to get one row per day with columns for each 15-minute interval\n",
" pivot_table = interval_sum.pivot(index='date', columns='time', values='value').fillna(0)\n",
"\n",
" # Step 8: Create a full range of 15-minute intervals (00:00:00 to 23:45:00)\n",
" full_time_range = pd.date_range('00:00', '23:45', freq='15T').time\n",
"\n",
" # Step 9: Reindex to include all possible 15-minute intervals and fill missing values with 0\n",
" pivot_table = pivot_table.reindex(columns=full_time_range, fill_value=0)\n",
"\n",
" # Step 10: Rename columns to reflect 15-minute intervals\n",
" pivot_table.columns = [f'{str(col)}' for col in pivot_table.columns]\n",
"\n",
" # Step 11: Reset index to have 'date' as a column instead of an index\n",
" pivot_table.reset_index(inplace=True)\n",
"\n",
" # Step 12: Add day of the week, month, and year columns\n",
" pivot_table['DayOfWeek'] = pd.to_datetime(pivot_table['date']).dt.day_name()\n",
" pivot_table['Month'] = pd.to_datetime(pivot_table['date']).dt.month\n",
" pivot_table['Year'] = pd.to_datetime(pivot_table['date']).dt.year\n",
"\n",
" # Step 13: One-hot encode the 'DayOfWeek' column\n",
" pivot_table = pd.concat([pivot_table, pd.get_dummies(pivot_table['DayOfWeek'], prefix='DayOfWeek')], axis=1)\n",
"\n",
" # Step 14: Convert 15-minute interval values to binary (True if > 0, else False)\n",
" for col in pivot_table.columns[1:97]: # Skip the 'date' column and focus on 15-minute intervals\n",
" pivot_table[col] = pivot_table[col].apply(lambda x: True if x > 0 else False)\n",
"\n",
" # Step 15: Add 'user' column with the specified user label\n",
" pivot_table['user'] = user_label\n",
"\n",
" # Print which file is currently being processed\n",
" print(f\"Processing file: {file_path}, User label: {user_label}\")\n",
"\n",
" # Step 16: Drop the 'DayOfWeek' column as it has been one-hot encoded\n",
" pivot_table.drop(columns=['DayOfWeek'], inplace=True)\n",
"\n",
" return pivot_table\n",
"\n",
"# List of files to skip\n",
"files_to_skip = {'StepCount06.csv','StepCount10.csv','StepCount12.csv', 'StepCount13.csv', 'StepCount15.csv', 'StepCount17.csv',\n",
" 'StepCount18.csv', 'StepCount20.csv', 'StepCount24.csv', 'StepCount27.csv','StepCount31.csv','StepCount32.csv',\n",
" 'StepCount42.csv', 'StepCount46.csv'}\n",
"\n",
"# Generate file paths, skipping specified files\n",
"file_paths = [f'/content/drive/My Drive/Data/iOS/StepCount{i:02d}.csv' for i in range(1, 47)\n",
" if f'StepCount{i:02d}.csv' not in files_to_skip]\n",
"\n",
"# Generate user labels based on file index\n",
"user_labels = list(range(len(file_paths)))\n",
"\n",
"# Process each file with its corresponding user label and concatenate the results\n",
"processed_dfs = [process_file(file_path, user_label) for file_path, user_label in zip(file_paths, user_labels)]\n",
"combined_df = pd.concat(processed_dfs, ignore_index=True)\n",
"\n",
"# Save the combined DataFrame to a new Excel file\n",
"updated_file_path = '/content/combined_aggregated_data_15min_with_threshold.xlsx'\n",
"combined_df.to_excel(updated_file_path, index=False)\n",
"\n",
"# Print the final DataFrame\n",
"print(combined_df)\n"
],
"metadata": {
"id": "w9RRwHj2wcbI",
"colab": {
"base_uri": "https://localhost:8080/"
},
"outputId": "da82f9b0-54f0-45eb-c823-109d1e244002",
"collapsed": true
},
"execution_count": null,
"outputs": [
{
"output_type": "stream",
"name": "stderr",
"text": [
"<ipython-input-7-a477c39a373a>:14: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['startDate'] = pd.to_datetime(iphone_df['startDate'], format='%Y-%m-%d %H:%M:%S %z')\n",
"<ipython-input-7-a477c39a373a>:17: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['15min_interval'] = iphone_df['startDate'].dt.floor('15T')\n",
"<ipython-input-7-a477c39a373a>:20: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['date'] = iphone_df['15min_interval'].dt.date\n",
"<ipython-input-7-a477c39a373a>:21: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['time'] = iphone_df['15min_interval'].dt.time\n"
]
},
{
"output_type": "stream",
"name": "stdout",
"text": [
"Processing file: /content/drive/My Drive/Data/iOS/StepCount01.csv, User label: 0\n"
]
},
{
"output_type": "stream",
"name": "stderr",
"text": [
"<ipython-input-7-a477c39a373a>:14: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['startDate'] = pd.to_datetime(iphone_df['startDate'], format='%Y-%m-%d %H:%M:%S %z')\n",
"<ipython-input-7-a477c39a373a>:17: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['15min_interval'] = iphone_df['startDate'].dt.floor('15T')\n",
"<ipython-input-7-a477c39a373a>:20: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['date'] = iphone_df['15min_interval'].dt.date\n",
"<ipython-input-7-a477c39a373a>:21: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['time'] = iphone_df['15min_interval'].dt.time\n"
]
},
{
"output_type": "stream",
"name": "stdout",
"text": [
"Processing file: /content/drive/My Drive/Data/iOS/StepCount02.csv, User label: 1\n",
"Processing file: /content/drive/My Drive/Data/iOS/StepCount03.csv, User label: 2\n",
"Processing file: /content/drive/My Drive/Data/iOS/StepCount04.csv, User label: 3\n"
]
},
{
"output_type": "stream",
"name": "stderr",
"text": [
"<ipython-input-7-a477c39a373a>:14: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['startDate'] = pd.to_datetime(iphone_df['startDate'], format='%Y-%m-%d %H:%M:%S %z')\n",
"<ipython-input-7-a477c39a373a>:17: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['15min_interval'] = iphone_df['startDate'].dt.floor('15T')\n",
"<ipython-input-7-a477c39a373a>:20: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['date'] = iphone_df['15min_interval'].dt.date\n",
"<ipython-input-7-a477c39a373a>:21: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['time'] = iphone_df['15min_interval'].dt.time\n"
]
},
{
"output_type": "stream",
"name": "stdout",
"text": [
"Processing file: /content/drive/My Drive/Data/iOS/StepCount05.csv, User label: 4\n",
"Processing file: /content/drive/My Drive/Data/iOS/StepCount07.csv, User label: 5\n"
]
},
{
"output_type": "stream",
"name": "stderr",
"text": [
"<ipython-input-7-a477c39a373a>:14: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['startDate'] = pd.to_datetime(iphone_df['startDate'], format='%Y-%m-%d %H:%M:%S %z')\n",
"<ipython-input-7-a477c39a373a>:17: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['15min_interval'] = iphone_df['startDate'].dt.floor('15T')\n",
"<ipython-input-7-a477c39a373a>:20: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['date'] = iphone_df['15min_interval'].dt.date\n",
"<ipython-input-7-a477c39a373a>:21: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['time'] = iphone_df['15min_interval'].dt.time\n",
"<ipython-input-7-a477c39a373a>:14: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['startDate'] = pd.to_datetime(iphone_df['startDate'], format='%Y-%m-%d %H:%M:%S %z')\n",
"<ipython-input-7-a477c39a373a>:17: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['15min_interval'] = iphone_df['startDate'].dt.floor('15T')\n",
"<ipython-input-7-a477c39a373a>:20: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['date'] = iphone_df['15min_interval'].dt.date\n",
"<ipython-input-7-a477c39a373a>:21: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['time'] = iphone_df['15min_interval'].dt.time\n"
]
},
{
"output_type": "stream",
"name": "stdout",
"text": [
"Processing file: /content/drive/My Drive/Data/iOS/StepCount08.csv, User label: 6\n",
"Processing file: /content/drive/My Drive/Data/iOS/StepCount09.csv, User label: 7\n",
"Processing file: /content/drive/My Drive/Data/iOS/StepCount11.csv, User label: 8\n",
"Processing file: /content/drive/My Drive/Data/iOS/StepCount14.csv, User label: 9\n",
"Processing file: /content/drive/My Drive/Data/iOS/StepCount16.csv, User label: 10\n",
"Processing file: /content/drive/My Drive/Data/iOS/StepCount19.csv, User label: 11\n",
"Processing file: /content/drive/My Drive/Data/iOS/StepCount21.csv, User label: 12\n",
"Processing file: /content/drive/My Drive/Data/iOS/StepCount22.csv, User label: 13\n"
]
},
{
"output_type": "stream",
"name": "stderr",
"text": [
"<ipython-input-7-a477c39a373a>:14: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['startDate'] = pd.to_datetime(iphone_df['startDate'], format='%Y-%m-%d %H:%M:%S %z')\n",
"<ipython-input-7-a477c39a373a>:17: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['15min_interval'] = iphone_df['startDate'].dt.floor('15T')\n",
"<ipython-input-7-a477c39a373a>:20: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['date'] = iphone_df['15min_interval'].dt.date\n",
"<ipython-input-7-a477c39a373a>:21: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['time'] = iphone_df['15min_interval'].dt.time\n"
]
},
{
"output_type": "stream",
"name": "stdout",
"text": [
"Processing file: /content/drive/My Drive/Data/iOS/StepCount23.csv, User label: 14\n"
]
},
{
"output_type": "stream",
"name": "stderr",
"text": [
"<ipython-input-7-a477c39a373a>:14: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['startDate'] = pd.to_datetime(iphone_df['startDate'], format='%Y-%m-%d %H:%M:%S %z')\n",
"<ipython-input-7-a477c39a373a>:17: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['15min_interval'] = iphone_df['startDate'].dt.floor('15T')\n",
"<ipython-input-7-a477c39a373a>:20: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['date'] = iphone_df['15min_interval'].dt.date\n",
"<ipython-input-7-a477c39a373a>:21: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['time'] = iphone_df['15min_interval'].dt.time\n"
]
},
{
"output_type": "stream",
"name": "stdout",
"text": [
"Processing file: /content/drive/My Drive/Data/iOS/StepCount25.csv, User label: 15\n",
"Processing file: /content/drive/My Drive/Data/iOS/StepCount26.csv, User label: 16\n"
]
},
{
"output_type": "stream",
"name": "stderr",
"text": [
"<ipython-input-7-a477c39a373a>:14: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['startDate'] = pd.to_datetime(iphone_df['startDate'], format='%Y-%m-%d %H:%M:%S %z')\n",
"<ipython-input-7-a477c39a373a>:17: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['15min_interval'] = iphone_df['startDate'].dt.floor('15T')\n",
"<ipython-input-7-a477c39a373a>:20: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['date'] = iphone_df['15min_interval'].dt.date\n",
"<ipython-input-7-a477c39a373a>:21: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['time'] = iphone_df['15min_interval'].dt.time\n"
]
},
{
"output_type": "stream",
"name": "stdout",
"text": [
"Processing file: /content/drive/My Drive/Data/iOS/StepCount28.csv, User label: 17\n",
"Processing file: /content/drive/My Drive/Data/iOS/StepCount29.csv, User label: 18\n",
"Processing file: /content/drive/My Drive/Data/iOS/StepCount30.csv, User label: 19\n"
]
},
{
"output_type": "stream",
"name": "stderr",
"text": [
"<ipython-input-7-a477c39a373a>:14: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['startDate'] = pd.to_datetime(iphone_df['startDate'], format='%Y-%m-%d %H:%M:%S %z')\n",
"<ipython-input-7-a477c39a373a>:17: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['15min_interval'] = iphone_df['startDate'].dt.floor('15T')\n",
"<ipython-input-7-a477c39a373a>:20: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['date'] = iphone_df['15min_interval'].dt.date\n",
"<ipython-input-7-a477c39a373a>:21: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['time'] = iphone_df['15min_interval'].dt.time\n"
]
},
{
"output_type": "stream",
"name": "stdout",
"text": [
"Processing file: /content/drive/My Drive/Data/iOS/StepCount33.csv, User label: 20\n",
"Processing file: /content/drive/My Drive/Data/iOS/StepCount34.csv, User label: 21\n",
"Processing file: /content/drive/My Drive/Data/iOS/StepCount35.csv, User label: 22\n",
"Processing file: /content/drive/My Drive/Data/iOS/StepCount36.csv, User label: 23\n"
]
},
{
"output_type": "stream",
"name": "stderr",
"text": [
"<ipython-input-7-a477c39a373a>:14: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['startDate'] = pd.to_datetime(iphone_df['startDate'], format='%Y-%m-%d %H:%M:%S %z')\n",
"<ipython-input-7-a477c39a373a>:17: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['15min_interval'] = iphone_df['startDate'].dt.floor('15T')\n",
"<ipython-input-7-a477c39a373a>:20: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['date'] = iphone_df['15min_interval'].dt.date\n",
"<ipython-input-7-a477c39a373a>:21: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['time'] = iphone_df['15min_interval'].dt.time\n"
]
},
{
"output_type": "stream",
"name": "stdout",
"text": [
"Processing file: /content/drive/My Drive/Data/iOS/StepCount37.csv, User label: 24\n"
]
},
{
"output_type": "stream",
"name": "stderr",
"text": [
"<ipython-input-7-a477c39a373a>:14: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['startDate'] = pd.to_datetime(iphone_df['startDate'], format='%Y-%m-%d %H:%M:%S %z')\n",
"<ipython-input-7-a477c39a373a>:17: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['15min_interval'] = iphone_df['startDate'].dt.floor('15T')\n",
"<ipython-input-7-a477c39a373a>:20: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['date'] = iphone_df['15min_interval'].dt.date\n",
"<ipython-input-7-a477c39a373a>:21: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['time'] = iphone_df['15min_interval'].dt.time\n"
]
},
{
"output_type": "stream",
"name": "stdout",
"text": [
"Processing file: /content/drive/My Drive/Data/iOS/StepCount38.csv, User label: 25\n",
"Processing file: /content/drive/My Drive/Data/iOS/StepCount39.csv, User label: 26\n"
]
},
{
"output_type": "stream",
"name": "stderr",
"text": [
"<ipython-input-7-a477c39a373a>:14: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['startDate'] = pd.to_datetime(iphone_df['startDate'], format='%Y-%m-%d %H:%M:%S %z')\n",
"<ipython-input-7-a477c39a373a>:17: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['15min_interval'] = iphone_df['startDate'].dt.floor('15T')\n",
"<ipython-input-7-a477c39a373a>:20: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['date'] = iphone_df['15min_interval'].dt.date\n",
"<ipython-input-7-a477c39a373a>:21: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['time'] = iphone_df['15min_interval'].dt.time\n"
]
},
{
"output_type": "stream",
"name": "stdout",
"text": [
"Processing file: /content/drive/My Drive/Data/iOS/StepCount40.csv, User label: 27\n"
]
},
{
"output_type": "stream",
"name": "stderr",
"text": [
"<ipython-input-7-a477c39a373a>:14: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['startDate'] = pd.to_datetime(iphone_df['startDate'], format='%Y-%m-%d %H:%M:%S %z')\n",
"<ipython-input-7-a477c39a373a>:17: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['15min_interval'] = iphone_df['startDate'].dt.floor('15T')\n",
"<ipython-input-7-a477c39a373a>:20: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['date'] = iphone_df['15min_interval'].dt.date\n",
"<ipython-input-7-a477c39a373a>:21: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['time'] = iphone_df['15min_interval'].dt.time\n"
]
},
{
"output_type": "stream",
"name": "stdout",
"text": [
"Processing file: /content/drive/My Drive/Data/iOS/StepCount41.csv, User label: 28\n",
"Processing file: /content/drive/My Drive/Data/iOS/StepCount43.csv, User label: 29\n",
"Processing file: /content/drive/My Drive/Data/iOS/StepCount44.csv, User label: 30\n",
"Processing file: /content/drive/My Drive/Data/iOS/StepCount45.csv, User label: 31\n",
" date 00:00:00 00:15:00 00:30:00 00:45:00 01:00:00 01:15:00 \\\n",
"0 2017-07-20 False False False False False False \n",
"1 2017-07-21 False False False False False False \n",
"2 2017-07-22 False False False True True False \n",
"3 2017-07-23 False False False False False False \n",
"4 2017-07-24 False True False False False False \n",
"... ... ... ... ... ... ... ... \n",
"36480 2020-06-09 False False True False False False \n",
"36481 2020-06-10 False False False False False False \n",
"36482 2020-06-11 False True False False False False \n",
"36483 2020-06-12 False False True False False False \n",
"36484 2020-06-13 False False True False False False \n",
"\n",
" 01:30:00 01:45:00 02:00:00 ... Month Year DayOfWeek_Friday \\\n",
"0 False False False ... 7 2017 False \n",
"1 False False False ... 7 2017 True \n",
"2 False False False ... 7 2017 False \n",
"3 False False False ... 7 2017 False \n",
"4 True False False ... 7 2017 False \n",
"... ... ... ... ... ... ... ... \n",
"36480 False False False ... 6 2020 False \n",
"36481 False False False ... 6 2020 False \n",
"36482 False False False ... 6 2020 False \n",
"36483 False False False ... 6 2020 True \n",
"36484 False False False ... 6 2020 False \n",
"\n",
" DayOfWeek_Monday DayOfWeek_Saturday DayOfWeek_Sunday \\\n",
"0 False False False \n",
"1 False False False \n",
"2 False True False \n",
"3 False False True \n",
"4 True False False \n",
"... ... ... ... \n",
"36480 False False False \n",
"36481 False False False \n",
"36482 False False False \n",
"36483 False False False \n",
"36484 False True False \n",
"\n",
" DayOfWeek_Thursday DayOfWeek_Tuesday DayOfWeek_Wednesday user \n",
"0 True False False 0 \n",
"1 False False False 0 \n",
"2 False False False 0 \n",
"3 False False False 0 \n",
"4 False False False 0 \n",
"... ... ... ... ... \n",
"36480 False True False 31 \n",
"36481 False False True 31 \n",
"36482 True False False 31 \n",
"36483 False False False 31 \n",
"36484 False False False 31 \n",
"\n",
"[36485 rows x 107 columns]\n"
]
}
]
},
{
"cell_type": "markdown",
"source": [
"FINAL CODE OF 1HR WITH THRESHOLD"
],
"metadata": {
"id": "Sab7rXWDKuvJ"
}
},
{
"cell_type": "code",
"source": [
"import pandas as pd\n",
"\n",
"def process_file(file_path, user_label):\n",
" # Load the dataset\n",
" df = pd.read_csv(file_path, delimiter=';')\n",
"\n",
" # Step 1: Filter for iPhone devices\n",
" iphone_df = df[df['device'].str.contains('iPhone', na=False)] # Treat NaN as False\n",
"\n",
" # Step 2: Select the desired columns\n",
" result = iphone_df[['startDate', 'endDate', 'value']]\n",
"\n",
" # Step 3: Convert startDate to datetime\n",
" iphone_df['startDate'] = pd.to_datetime(iphone_df['startDate'], format='%Y-%m-%d %H:%M:%S %z')\n",
"\n",
" # Step 4: Round down the startDate to the nearest 1-hour interval\n",
" iphone_df['1hr_interval'] = iphone_df['startDate'].dt.floor('H')\n",
"\n",
" # Step 5: Extract date and time\n",
" iphone_df['date'] = iphone_df['1hr_interval'].dt.date\n",
" iphone_df['time'] = iphone_df['1hr_interval'].dt.time\n",
"\n",
" # Step 6: Group by date and time, then sum the values for 1-hour intervals\n",
" iphone_df_filtered = iphone_df[iphone_df['value'] > 25].dropna(subset=['value'])\n",
" interval_sum = iphone_df.groupby(['date', 'time'])['value'].sum().reset_index()\n",
"\n",
" # Step 7: Pivot the data to get one row per day with columns for each 1-hour interval\n",
" pivot_table = interval_sum.pivot(index='date', columns='time', values='value').fillna(0)\n",
"\n",
" # Step 8: Create a full range of 1-hour intervals (00:00:00 to 23:00:00)\n",
" full_time_range = pd.date_range('00:00', '23:00', freq='H').time\n",
"\n",
" # Step 9: Reindex to include all possible 1-hour intervals and fill missing values with 0\n",
" pivot_table = pivot_table.reindex(columns=full_time_range, fill_value=0)\n",
"\n",
" # Step 10: Rename columns to reflect 1-hour intervals\n",
" pivot_table.columns = [f'{str(col)}' for col in pivot_table.columns]\n",
"\n",
" # Step 11: Reset index to have 'date' as a column instead of an index\n",
" pivot_table.reset_index(inplace=True)\n",
"\n",
" # Step 12: Add day of the week, month, and year columns\n",
" pivot_table['DayOfWeek'] = pd.to_datetime(pivot_table['date']).dt.day_name()\n",
" pivot_table['Month'] = pd.to_datetime(pivot_table['date']).dt.month\n",
" pivot_table['Year'] = pd.to_datetime(pivot_table['date']).dt.year\n",
"\n",
" # Step 13: One-hot encode the 'DayOfWeek' column\n",
" pivot_table = pd.concat([pivot_table, pd.get_dummies(pivot_table['DayOfWeek'], prefix='DayOfWeek')], axis=1)\n",
"\n",
" # Step 14: Convert 1-hour interval values to binary (True if > 0, else False)\n",
" for col in pivot_table.columns[1:25]: # Skip the 'date' column and focus on 1-hour intervals\n",
" pivot_table[col] = pivot_table[col].apply(lambda x: True if x > 0 else False)\n",
"\n",
" # Step 15: Add 'user' column with the specified user label\n",
" pivot_table['user'] = user_label\n",
"\n",
" # Print which file is currently being processed\n",
" print(f\"Processing file: {file_path}, User label: {user_label}\")\n",
"\n",
" # Step 16: Drop the 'DayOfWeek' column as it has been one-hot encoded\n",
" pivot_table.drop(columns=['DayOfWeek'], inplace=True)\n",
"\n",
" return pivot_table\n",
"\n",
"# List of files to skip\n",
"files_to_skip = {'StepCount06.csv','StepCount10.csv','StepCount12.csv', 'StepCount13.csv', 'StepCount15.csv', 'StepCount17.csv',\n",
" 'StepCount18.csv', 'StepCount20.csv', 'StepCount24.csv', 'StepCount27.csv','StepCount31.csv','StepCount32.csv',\n",
" 'StepCount42.csv', 'StepCount46.csv'}\n",
"\n",
"# Generate file paths, skipping specified files\n",
"file_paths = [f'/content/drive/My Drive/Data/iOS/StepCount{i:02d}.csv' for i in range(1, 47)\n",
" if f'StepCount{i:02d}.csv' not in files_to_skip]\n",
"\n",
"# Generate user labels based on file index\n",
"user_labels = list(range(len(file_paths)))\n",
"\n",
"# Process each file with its corresponding user label and concatenate the results\n",
"processed_dfs = [process_file(file_path, user_label) for file_path, user_label in zip(file_paths, user_labels)]\n",
"combined_df = pd.concat(processed_dfs, ignore_index=True)\n",
"\n",
"# Save the combined DataFrame to a new Excel file\n",
"updated_file_path = '/content/combined_aggregated_data_1hr_withthreshold.xlsx'\n",
"combined_df.to_excel(updated_file_path, index=False)\n",
"\n",
"# Print the final DataFrame\n",
"print(combined_df)\n"
],
"metadata": {
"id": "ovDzbhzW6GQX",
"colab": {
"base_uri": "https://localhost:8080/"
},
"outputId": "024b89fb-6990-4b8c-b69f-9bfa27f0ce5e",
"collapsed": true
},
"execution_count": null,
"outputs": [
{
"output_type": "stream",
"name": "stderr",
"text": [
"<ipython-input-8-ae3c61a191b5>:14: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['startDate'] = pd.to_datetime(iphone_df['startDate'], format='%Y-%m-%d %H:%M:%S %z')\n",
"<ipython-input-8-ae3c61a191b5>:17: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['1hr_interval'] = iphone_df['startDate'].dt.floor('H')\n",
"<ipython-input-8-ae3c61a191b5>:20: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['date'] = iphone_df['1hr_interval'].dt.date\n",
"<ipython-input-8-ae3c61a191b5>:21: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['time'] = iphone_df['1hr_interval'].dt.time\n"
]
},
{
"output_type": "stream",
"name": "stdout",
"text": [
"Processing file: /content/drive/My Drive/Data/iOS/StepCount01.csv, User label: 0\n"
]
},
{
"output_type": "stream",
"name": "stderr",
"text": [
"<ipython-input-8-ae3c61a191b5>:14: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['startDate'] = pd.to_datetime(iphone_df['startDate'], format='%Y-%m-%d %H:%M:%S %z')\n",
"<ipython-input-8-ae3c61a191b5>:17: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['1hr_interval'] = iphone_df['startDate'].dt.floor('H')\n",
"<ipython-input-8-ae3c61a191b5>:20: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['date'] = iphone_df['1hr_interval'].dt.date\n",
"<ipython-input-8-ae3c61a191b5>:21: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['time'] = iphone_df['1hr_interval'].dt.time\n"
]
},
{
"output_type": "stream",
"name": "stdout",
"text": [
"Processing file: /content/drive/My Drive/Data/iOS/StepCount02.csv, User label: 1\n",
"Processing file: /content/drive/My Drive/Data/iOS/StepCount03.csv, User label: 2\n",
"Processing file: /content/drive/My Drive/Data/iOS/StepCount04.csv, User label: 3\n"
]
},
{
"output_type": "stream",
"name": "stderr",
"text": [
"<ipython-input-8-ae3c61a191b5>:14: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['startDate'] = pd.to_datetime(iphone_df['startDate'], format='%Y-%m-%d %H:%M:%S %z')\n",
"<ipython-input-8-ae3c61a191b5>:17: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['1hr_interval'] = iphone_df['startDate'].dt.floor('H')\n",
"<ipython-input-8-ae3c61a191b5>:20: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['date'] = iphone_df['1hr_interval'].dt.date\n",
"<ipython-input-8-ae3c61a191b5>:21: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['time'] = iphone_df['1hr_interval'].dt.time\n"
]
},
{
"output_type": "stream",
"name": "stdout",
"text": [
"Processing file: /content/drive/My Drive/Data/iOS/StepCount05.csv, User label: 4\n",
"Processing file: /content/drive/My Drive/Data/iOS/StepCount07.csv, User label: 5\n"
]
},
{
"output_type": "stream",
"name": "stderr",
"text": [
"<ipython-input-8-ae3c61a191b5>:14: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['startDate'] = pd.to_datetime(iphone_df['startDate'], format='%Y-%m-%d %H:%M:%S %z')\n",
"<ipython-input-8-ae3c61a191b5>:17: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['1hr_interval'] = iphone_df['startDate'].dt.floor('H')\n",
"<ipython-input-8-ae3c61a191b5>:20: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['date'] = iphone_df['1hr_interval'].dt.date\n",
"<ipython-input-8-ae3c61a191b5>:21: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['time'] = iphone_df['1hr_interval'].dt.time\n",
"<ipython-input-8-ae3c61a191b5>:14: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['startDate'] = pd.to_datetime(iphone_df['startDate'], format='%Y-%m-%d %H:%M:%S %z')\n",
"<ipython-input-8-ae3c61a191b5>:17: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['1hr_interval'] = iphone_df['startDate'].dt.floor('H')\n",
"<ipython-input-8-ae3c61a191b5>:20: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['date'] = iphone_df['1hr_interval'].dt.date\n",
"<ipython-input-8-ae3c61a191b5>:21: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['time'] = iphone_df['1hr_interval'].dt.time\n"
]
},
{
"output_type": "stream",
"name": "stdout",
"text": [
"Processing file: /content/drive/My Drive/Data/iOS/StepCount08.csv, User label: 6\n",
"Processing file: /content/drive/My Drive/Data/iOS/StepCount09.csv, User label: 7\n",
"Processing file: /content/drive/My Drive/Data/iOS/StepCount11.csv, User label: 8\n",
"Processing file: /content/drive/My Drive/Data/iOS/StepCount14.csv, User label: 9\n",
"Processing file: /content/drive/My Drive/Data/iOS/StepCount16.csv, User label: 10\n",
"Processing file: /content/drive/My Drive/Data/iOS/StepCount19.csv, User label: 11\n",
"Processing file: /content/drive/My Drive/Data/iOS/StepCount21.csv, User label: 12\n",
"Processing file: /content/drive/My Drive/Data/iOS/StepCount22.csv, User label: 13\n"
]
},
{
"output_type": "stream",
"name": "stderr",
"text": [
"<ipython-input-8-ae3c61a191b5>:14: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['startDate'] = pd.to_datetime(iphone_df['startDate'], format='%Y-%m-%d %H:%M:%S %z')\n",
"<ipython-input-8-ae3c61a191b5>:17: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['1hr_interval'] = iphone_df['startDate'].dt.floor('H')\n",
"<ipython-input-8-ae3c61a191b5>:20: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['date'] = iphone_df['1hr_interval'].dt.date\n",
"<ipython-input-8-ae3c61a191b5>:21: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['time'] = iphone_df['1hr_interval'].dt.time\n"
]
},
{
"output_type": "stream",
"name": "stdout",
"text": [
"Processing file: /content/drive/My Drive/Data/iOS/StepCount23.csv, User label: 14\n"
]
},
{
"output_type": "stream",
"name": "stderr",
"text": [
"<ipython-input-8-ae3c61a191b5>:14: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['startDate'] = pd.to_datetime(iphone_df['startDate'], format='%Y-%m-%d %H:%M:%S %z')\n",
"<ipython-input-8-ae3c61a191b5>:17: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['1hr_interval'] = iphone_df['startDate'].dt.floor('H')\n",
"<ipython-input-8-ae3c61a191b5>:20: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['date'] = iphone_df['1hr_interval'].dt.date\n",
"<ipython-input-8-ae3c61a191b5>:21: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['time'] = iphone_df['1hr_interval'].dt.time\n"
]
},
{
"output_type": "stream",
"name": "stdout",
"text": [
"Processing file: /content/drive/My Drive/Data/iOS/StepCount25.csv, User label: 15\n",
"Processing file: /content/drive/My Drive/Data/iOS/StepCount26.csv, User label: 16\n"
]
},
{
"output_type": "stream",
"name": "stderr",
"text": [
"<ipython-input-8-ae3c61a191b5>:14: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['startDate'] = pd.to_datetime(iphone_df['startDate'], format='%Y-%m-%d %H:%M:%S %z')\n",
"<ipython-input-8-ae3c61a191b5>:17: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['1hr_interval'] = iphone_df['startDate'].dt.floor('H')\n",
"<ipython-input-8-ae3c61a191b5>:20: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['date'] = iphone_df['1hr_interval'].dt.date\n",
"<ipython-input-8-ae3c61a191b5>:21: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['time'] = iphone_df['1hr_interval'].dt.time\n"
]
},
{
"output_type": "stream",
"name": "stdout",
"text": [
"Processing file: /content/drive/My Drive/Data/iOS/StepCount28.csv, User label: 17\n",
"Processing file: /content/drive/My Drive/Data/iOS/StepCount29.csv, User label: 18\n",
"Processing file: /content/drive/My Drive/Data/iOS/StepCount30.csv, User label: 19\n"
]
},
{
"output_type": "stream",
"name": "stderr",
"text": [
"<ipython-input-8-ae3c61a191b5>:14: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['startDate'] = pd.to_datetime(iphone_df['startDate'], format='%Y-%m-%d %H:%M:%S %z')\n",
"<ipython-input-8-ae3c61a191b5>:17: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['1hr_interval'] = iphone_df['startDate'].dt.floor('H')\n",
"<ipython-input-8-ae3c61a191b5>:20: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['date'] = iphone_df['1hr_interval'].dt.date\n",
"<ipython-input-8-ae3c61a191b5>:21: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['time'] = iphone_df['1hr_interval'].dt.time\n"
]
},
{
"output_type": "stream",
"name": "stdout",
"text": [
"Processing file: /content/drive/My Drive/Data/iOS/StepCount33.csv, User label: 20\n",
"Processing file: /content/drive/My Drive/Data/iOS/StepCount34.csv, User label: 21\n",
"Processing file: /content/drive/My Drive/Data/iOS/StepCount35.csv, User label: 22\n",
"Processing file: /content/drive/My Drive/Data/iOS/StepCount36.csv, User label: 23\n"
]
},
{
"output_type": "stream",
"name": "stderr",
"text": [
"<ipython-input-8-ae3c61a191b5>:14: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['startDate'] = pd.to_datetime(iphone_df['startDate'], format='%Y-%m-%d %H:%M:%S %z')\n",
"<ipython-input-8-ae3c61a191b5>:17: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['1hr_interval'] = iphone_df['startDate'].dt.floor('H')\n",
"<ipython-input-8-ae3c61a191b5>:20: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['date'] = iphone_df['1hr_interval'].dt.date\n",
"<ipython-input-8-ae3c61a191b5>:21: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['time'] = iphone_df['1hr_interval'].dt.time\n"
]
},
{
"output_type": "stream",
"name": "stdout",
"text": [
"Processing file: /content/drive/My Drive/Data/iOS/StepCount37.csv, User label: 24\n"
]
},
{
"output_type": "stream",
"name": "stderr",
"text": [
"<ipython-input-8-ae3c61a191b5>:14: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['startDate'] = pd.to_datetime(iphone_df['startDate'], format='%Y-%m-%d %H:%M:%S %z')\n",
"<ipython-input-8-ae3c61a191b5>:17: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['1hr_interval'] = iphone_df['startDate'].dt.floor('H')\n",
"<ipython-input-8-ae3c61a191b5>:20: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['date'] = iphone_df['1hr_interval'].dt.date\n",
"<ipython-input-8-ae3c61a191b5>:21: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['time'] = iphone_df['1hr_interval'].dt.time\n"
]
},
{
"output_type": "stream",
"name": "stdout",
"text": [
"Processing file: /content/drive/My Drive/Data/iOS/StepCount38.csv, User label: 25\n",
"Processing file: /content/drive/My Drive/Data/iOS/StepCount39.csv, User label: 26\n"
]
},
{
"output_type": "stream",
"name": "stderr",
"text": [
"<ipython-input-8-ae3c61a191b5>:14: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['startDate'] = pd.to_datetime(iphone_df['startDate'], format='%Y-%m-%d %H:%M:%S %z')\n",
"<ipython-input-8-ae3c61a191b5>:17: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['1hr_interval'] = iphone_df['startDate'].dt.floor('H')\n",
"<ipython-input-8-ae3c61a191b5>:20: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['date'] = iphone_df['1hr_interval'].dt.date\n",
"<ipython-input-8-ae3c61a191b5>:21: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['time'] = iphone_df['1hr_interval'].dt.time\n"
]
},
{
"output_type": "stream",
"name": "stdout",
"text": [
"Processing file: /content/drive/My Drive/Data/iOS/StepCount40.csv, User label: 27\n"
]
},
{
"output_type": "stream",
"name": "stderr",
"text": [
"<ipython-input-8-ae3c61a191b5>:14: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['startDate'] = pd.to_datetime(iphone_df['startDate'], format='%Y-%m-%d %H:%M:%S %z')\n",
"<ipython-input-8-ae3c61a191b5>:17: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['1hr_interval'] = iphone_df['startDate'].dt.floor('H')\n",
"<ipython-input-8-ae3c61a191b5>:20: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['date'] = iphone_df['1hr_interval'].dt.date\n",
"<ipython-input-8-ae3c61a191b5>:21: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" iphone_df['time'] = iphone_df['1hr_interval'].dt.time\n"
]
},
{
"output_type": "stream",
"name": "stdout",
"text": [
"Processing file: /content/drive/My Drive/Data/iOS/StepCount41.csv, User label: 28\n",
"Processing file: /content/drive/My Drive/Data/iOS/StepCount43.csv, User label: 29\n",
"Processing file: /content/drive/My Drive/Data/iOS/StepCount44.csv, User label: 30\n",
"Processing file: /content/drive/My Drive/Data/iOS/StepCount45.csv, User label: 31\n",
" date 00:00:00 01:00:00 02:00:00 03:00:00 04:00:00 05:00:00 \\\n",
"0 2017-07-20 False False False False False False \n",
"1 2017-07-21 False False False False False False \n",
"2 2017-07-22 True True False False False False \n",
"3 2017-07-23 False False False False False False \n",
"4 2017-07-24 True True False False False False \n",
"... ... ... ... ... ... ... ... \n",
"36480 2020-06-09 True False False False False False \n",
"36481 2020-06-10 False False True False False False \n",
"36482 2020-06-11 True False False False False False \n",
"36483 2020-06-12 True False False False False False \n",
"36484 2020-06-13 True False False False False False \n",
"\n",
" 06:00:00 07:00:00 08:00:00 ... Month Year DayOfWeek_Friday \\\n",
"0 False False False ... 7 2017 False \n",
"1 False True True ... 7 2017 True \n",
"2 False True True ... 7 2017 False \n",
"3 False False True ... 7 2017 False \n",
"4 False False True ... 7 2017 False \n",
"... ... ... ... ... ... ... ... \n",
"36480 False False False ... 6 2020 False \n",
"36481 False False False ... 6 2020 False \n",
"36482 True True True ... 6 2020 False \n",
"36483 False False False ... 6 2020 True \n",
"36484 False False False ... 6 2020 False \n",
"\n",
" DayOfWeek_Monday DayOfWeek_Saturday DayOfWeek_Sunday \\\n",
"0 False False False \n",
"1 False False False \n",
"2 False True False \n",
"3 False False True \n",
"4 True False False \n",
"... ... ... ... \n",
"36480 False False False \n",
"36481 False False False \n",
"36482 False False False \n",
"36483 False False False \n",
"36484 False True False \n",
"\n",
" DayOfWeek_Thursday DayOfWeek_Tuesday DayOfWeek_Wednesday user \n",
"0 True False False 0 \n",
"1 False False False 0 \n",
"2 False False False 0 \n",
"3 False False False 0 \n",
"4 False False False 0 \n",
"... ... ... ... ... \n",
"36480 False True False 31 \n",
"36481 False False True 31 \n",
"36482 True False False 31 \n",
"36483 False False False 31 \n",
"36484 False False False 31 \n",
"\n",
"[36485 rows x 35 columns]\n"
]
}
]
},
{
"cell_type": "code",
"source": [],
"metadata": {
"id": "6MLHoHlckLYk"
},
"execution_count": null,
"outputs": []
}
]
}