ChatGPT Code InterpreterでFitbit睡眠ログを可視化してみた

#!/usr/bin/env python3

import fitbit
import datetime
from oauthlib.oauth2 import LegacyApplicationClient
from requests_oauthlib import OAuth2Session

# アプリケーションの設定
client_id = ''
client_secret = ''
user_id = ''

# ユーザーの設定
access_token = ''
refresh_token = ''

# OAuth2セッションを作成
auth = OAuth2Session(client_id, token={
    'access_token': access_token,
    'refresh_token': refresh_token,
    'token_type': 'Bearer',
    'expires_in': '-30',
    'expires_at': '-30'
}, client=LegacyApplicationClient(client_id))

# Fitbitクライアントを作成
client = fitbit.Fitbit(client_id,
                       client_secret,
                       oauth2=True,
                       access_token=access_token,
                       refresh_token=refresh_token,
                       system='en_GB')

# 睡眠データを取得
start_date = datetime.date(2023, 7, 30)
end_date = start_date + datetime.timedelta(days=7)
sleep = client.get_sleep(date=start_date)

# 睡眠データを表示
print(sleep)

import matplotlib.pyplot as plt

# 睡眠データを取得
start_date = datetime.date(2023, 7, 14)
end_date = start_date + datetime.timedelta(days=14)

dates = []
total_sleep_times = []

for day in (start_date + datetime.timedelta(n) for n in range((end_date-start_date).days)):
    sleep = client.get_sleep(date=day)
    dates.append(day)
    total_sleep_time = sum([sl['duration'] for sl in sleep['sleep']]) / 3600000  # ms to hours
    total_sleep_times.append(total_sleep_time)

# 睡眠データを可視化
plt.figure(figsize=(10, 5))
plt.plot(dates, total_sleep_times)
plt.xlabel('Date')
plt.ylabel('Total Sleep Time (hours)')
plt.title('Sleep Time over Two Weeks')
plt.show()

import json

file_path = "/home/tsu-nera/Downloads/Takeout/Fitbit/Global Export Data/sleep-2023-07-15.json"

# Load the json file
with open(file_path) as file:
    data = json.load(file)

# Create a DataFrame for the sleep duration
duration_df = pd.DataFrame({'dateOfSleep': [record['dateOfSleep'] for record in data],
                            'minutesAsleep': [record['minutesAsleep'] for record in data]})

# Convert string date to datetime
duration_df['dateOfSleep'] = pd.to_datetime(duration_df['dateOfSleep'])

# Sort by date
duration_df = duration_df.sort_values('dateOfSleep')


# Add columns for each sleep stage duration
duration_df['deep'] = [record['levels']['summary']['deep']['minutes'] for record in data]
duration_df['wake'] = [record['levels']['summary']['wake']['minutes'] for record in data]
duration_df['light'] = [record['levels']['summary']['light']['minutes'] for record in data]
duration_df['rem'] = [record['levels']['summary']['rem']['minutes'] for record in data]

# Plot sleep duration by date with sleep stage proportions
plt.figure(figsize=(10, 6))
plt.bar(duration_df['dateOfSleep'], duration_df['deep'], label='Deep', color='#5A9')
plt.bar(duration_df['dateOfSleep'], duration_df['wake'], bottom=duration_df['deep'], label='Wake', color='#933')
plt.bar(duration_df['dateOfSleep'], duration_df['light'], bottom=duration_df['deep'] + duration_df['wake'], label='Light', color='#A77')
plt.bar(duration_df['dateOfSleep'], duration_df['rem'], bottom=duration_df['deep'] + duration_df['wake'] + duration_df['light'], label='REM', color='#39F')
plt.title('Sleep Duration by Date with Sleep Stages')
plt.xlabel('Date')
plt.ylabel('Sleep Duration (minutes)')
plt.gca().xaxis.set_major_formatter(mdates.DateFormatter('%Y-%m-%d'))  # format x-axis label to year-month-day
plt.gca().xaxis.set_major_locator(mdates.DayLocator())
plt.gcf().autofmt_xdate()  # rotate x-axis labels
plt.legend(loc='upper left')
plt.show()