ThonyPrice

z = df.copy(deep=True)
z = z.join(pd.get_dummies(z['Status']), how='left')

day_map = ['Mon', 'Tues', 'Wed', 'Thurs', 'Fri', 'Sat', 'Sun']
z['weekday'] = z.index.weekday
z = z[['weekday', 'glucose']]
z = z.resample('D').mean()
z['chunk'] = 'Second'
z['chunk'].iloc[:(z.shape[0]//2)] = 'First'
z.sort_values(['weekday', 'chunk'], inplace=True)

ThonyPrice

z = df.copy(deep=True)
z['day'] = z.index.date
z['day_time'] = z.index.round('5Min').time
z = z[['day', 'day_time', 'glucose']]
z = z.groupby(['day', 'day_time']).last()
z = z.reset_index()

_ = plt.figure(figsize=(20, 8))
x_cmap = sns.diverging_palette(220, 20, n=9)
f = sns.heatmap(z.pivot('day_time', 'day', 'glucose'), cmap=x_cmap, vmin=0, vmax=300)

ThonyPrice

z = df.copy(deep=True)
z = z.join(pd.get_dummies(z['Status']), how='left')

day_map = ['Mon', 'Tues', 'Wed', 'Thurs', 'Fri', 'Sat', 'Sun']
z['weekday'] = z.index.weekday
z = z[['weekday', 'glucose']]
z = z.resample('D').mean()
z.sort_values('weekday', inplace=True)
z.dropna(inplace=True)
z['weekday'] = z['weekday'].apply(lambda x: day_map[int(x)])

ThonyPrice

z = df.copy(deep=True)
z.index = z.index.round('5Min')

plt.figure(figsize=(16, 6))
ax = sns.lineplot(x=z.index.time, y=z['glucose'], color='#13ac5f')

ThonyPrice

def applyThreshold(val, urgentLow=54, low=70, high=180):
  if val > high:
      return 'high'
  elif val <= high and val >= low:
    return 'in_range'
  elif val < low and val > urgentLow:
    return 'low'
  else:
    return 'urgent_low';

ThonyPrice

z = df.copy(deep=True)
z = z.resample('D').mean()
z['date'] = z.index.date
z.insert(z.shape[1], 'day', z.index.value_counts().sort_index().cumsum())

plt.figure(figsize=(16, 6))
fs = sns.regplot(
  data = z, 
  x ='day', 
  y ='glucose', 

ThonyPrice

df['glucose'] = df['glucose'].astype(int)
_ = sns.distplot(df['glucose'])
df.describe()

ThonyPrice

expected_diff = timedelta(minutes=5, seconds=30)
missing_samples_idxs = np.where(z['t_diff'] > expected_diff)[0]

c_samples_times = z.iloc[missing_samples_idxs, :]['t_diff']
c_samples_times

ThonyPrice

# Inspect samples closer than average
import numpy as np
import matplotlib.pyplot as plt

expected_diff = timedelta(minutes=4, seconds=30)
close_samples_idxs = np.where(z['t_diff'] < expected_diff)[0]

c_samples_times = z.iloc[close_samples_idxs, :]['t_diff']
c_samples_times = c_samples_times.apply(lambda x: x.seconds)

ThonyPrice

from datetime import datetime, date

z = df.copy(deep=True)
z['t1'] = z.index
z['t2'] = z['t1'].shift(1)
z['t_diff'] = z['t1'] - z['t2']
z['t_diff'].describe()