Table of contents¶
- Introduction
- Processing
2.1 Loading data
2.2 Verifying loaded data
2.3 Missing values
2.4 Data type conversion
2.5 Duplicated values - Analysis
3.1 Daily activity tracking
3.2 Daily activity intensity
3.3 Sleep activity tracking
3.4 Weight information tracking
3.5 Minutes of average total daily activity
3.6 Hourly steps distribution
3.7 Activity intensity distribution - Conclusions
Introduction ¶
Bellabeat is a high-tech company that manufactures health-focused smart products and Urska Srsen, Bellabeat's cofounder, asked to analyze smart device usage data in order to gain insight on how people already use smart devices. By using this information the company is looking for recommendations that would help Bellabeat marketing strategy.
The data available for the analysis has been generated thanks to Amazon Mechanical Turk.
Data preprocessing ¶
In [64]:
import pandas as pd
import matplotlib.pyplot as plt
In [65]:
from google.colab import drive
drive.mount('/content/drive')
Drive already mounted at /content/drive; to attempt to forcibly remount, call drive.mount("/content/drive", force_remount=True).
Loading data ¶
In [66]:
# Datasets are being imported
# Notes: - "dailyActivity" has the same information as "dailySteps", "dailyIntensities" and "dailyCalories"
# so it's not needed to import the other three datasets
dailyActivity = pd.read_csv("/content/drive/MyDrive/Data Analysis /Google Data Analytics certificate/Fitabase Data 4.12.16-5.12.16/dailyActivity.csv")
hourlyIntensities = pd.read_csv("/content/drive/MyDrive/Data Analysis /Google Data Analytics certificate/Fitabase Data 4.12.16-5.12.16/hourlyIntensities.csv")
hourlySteps = pd.read_csv("/content/drive/MyDrive/Data Analysis /Google Data Analytics certificate/Fitabase Data 4.12.16-5.12.16/hourlySteps.csv")
sleepDay = pd.read_csv("/content/drive/MyDrive/Data Analysis /Google Data Analytics certificate/Fitabase Data 4.12.16-5.12.16/sleepDay.csv")
weightLogInfo = pd.read_csv("/content/drive/MyDrive/Data Analysis /Google Data Analytics certificate/Fitabase Data 4.12.16-5.12.16/weightLogInfo.csv")
Verifying loaded data ¶
In [67]:
# Checking if datasets have been imported correctly
dailyActivity.head(5)
Out[67]:
| Id | ActivityDate | TotalSteps | TotalDistance | TrackerDistance | LoggedActivitiesDistance | VeryActiveDistance | ModeratelyActiveDistance | LightActiveDistance | SedentaryActiveDistance | VeryActiveMinutes | FairlyActiveMinutes | LightlyActiveMinutes | SedentaryMinutes | Calories | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 0 | 1503960366 | 4/12/2016 | 13162 | 8.50 | 8.50 | 0.0 | 1.88 | 0.55 | 6.06 | 0.0 | 25 | 13 | 328 | 728 | 1985 |
| 1 | 1503960366 | 4/13/2016 | 10735 | 6.97 | 6.97 | 0.0 | 1.57 | 0.69 | 4.71 | 0.0 | 21 | 19 | 217 | 776 | 1797 |
| 2 | 1503960366 | 4/14/2016 | 10460 | 6.74 | 6.74 | 0.0 | 2.44 | 0.40 | 3.91 | 0.0 | 30 | 11 | 181 | 1218 | 1776 |
| 3 | 1503960366 | 4/15/2016 | 9762 | 6.28 | 6.28 | 0.0 | 2.14 | 1.26 | 2.83 | 0.0 | 29 | 34 | 209 | 726 | 1745 |
| 4 | 1503960366 | 4/16/2016 | 12669 | 8.16 | 8.16 | 0.0 | 2.71 | 0.41 | 5.04 | 0.0 | 36 | 10 | 221 | 773 | 1863 |
In [68]:
hourlyIntensities.head(5)
Out[68]:
| Id | ActivityHour | TotalIntensity | AverageIntensity | |
|---|---|---|---|---|
| 0 | 1503960366 | 4/12/2016 12:00:00 AM | 20 | 0.333333 |
| 1 | 1503960366 | 4/12/2016 1:00:00 AM | 8 | 0.133333 |
| 2 | 1503960366 | 4/12/2016 2:00:00 AM | 7 | 0.116667 |
| 3 | 1503960366 | 4/12/2016 3:00:00 AM | 0 | 0.000000 |
| 4 | 1503960366 | 4/12/2016 4:00:00 AM | 0 | 0.000000 |
In [69]:
hourlySteps.head(5)
Out[69]:
| Id | ActivityHour | StepTotal | |
|---|---|---|---|
| 0 | 1503960366 | 4/12/2016 12:00:00 AM | 373 |
| 1 | 1503960366 | 4/12/2016 1:00:00 AM | 160 |
| 2 | 1503960366 | 4/12/2016 2:00:00 AM | 151 |
| 3 | 1503960366 | 4/12/2016 3:00:00 AM | 0 |
| 4 | 1503960366 | 4/12/2016 4:00:00 AM | 0 |
In [70]:
sleepDay.head(5)
Out[70]:
| Id | SleepDay | TotalSleepRecords | TotalMinutesAsleep | TotalTimeInBed | |
|---|---|---|---|---|---|
| 0 | 1503960366 | 4/12/2016 12:00:00 AM | 1 | 327 | 346 |
| 1 | 1503960366 | 4/13/2016 12:00:00 AM | 2 | 384 | 407 |
| 2 | 1503960366 | 4/15/2016 12:00:00 AM | 1 | 412 | 442 |
| 3 | 1503960366 | 4/16/2016 12:00:00 AM | 2 | 340 | 367 |
| 4 | 1503960366 | 4/17/2016 12:00:00 AM | 1 | 700 | 712 |
In [71]:
weightLogInfo.head(5)
Out[71]:
| Id | Date | WeightKg | WeightPounds | Fat | BMI | IsManualReport | LogId | |
|---|---|---|---|---|---|---|---|---|
| 0 | 1503960366 | 5/2/2016 11:59:59 PM | 52.599998 | 115.963147 | 22.0 | 22.650000 | True | 1462233599000 |
| 1 | 1503960366 | 5/3/2016 11:59:59 PM | 52.599998 | 115.963147 | NaN | 22.650000 | True | 1462319999000 |
| 2 | 1927972279 | 4/13/2016 1:08:52 AM | 133.500000 | 294.317120 | NaN | 47.540001 | False | 1460509732000 |
| 3 | 2873212765 | 4/21/2016 11:59:59 PM | 56.700001 | 125.002104 | NaN | 21.450001 | True | 1461283199000 |
| 4 | 2873212765 | 5/12/2016 11:59:59 PM | 57.299999 | 126.324875 | NaN | 21.690001 | True | 1463097599000 |
Missing values ¶
In [72]:
# Checking for missing values and correct datatypes
dailyActivity.info()
<class 'pandas.core.frame.DataFrame'> RangeIndex: 940 entries, 0 to 939 Data columns (total 15 columns): # Column Non-Null Count Dtype --- ------ -------------- ----- 0 Id 940 non-null int64 1 ActivityDate 940 non-null object 2 TotalSteps 940 non-null int64 3 TotalDistance 940 non-null float64 4 TrackerDistance 940 non-null float64 5 LoggedActivitiesDistance 940 non-null float64 6 VeryActiveDistance 940 non-null float64 7 ModeratelyActiveDistance 940 non-null float64 8 LightActiveDistance 940 non-null float64 9 SedentaryActiveDistance 940 non-null float64 10 VeryActiveMinutes 940 non-null int64 11 FairlyActiveMinutes 940 non-null int64 12 LightlyActiveMinutes 940 non-null int64 13 SedentaryMinutes 940 non-null int64 14 Calories 940 non-null int64 dtypes: float64(7), int64(7), object(1) memory usage: 110.3+ KB
In [73]:
hourlyIntensities.info()
<class 'pandas.core.frame.DataFrame'> RangeIndex: 22099 entries, 0 to 22098 Data columns (total 4 columns): # Column Non-Null Count Dtype --- ------ -------------- ----- 0 Id 22099 non-null int64 1 ActivityHour 22099 non-null object 2 TotalIntensity 22099 non-null int64 3 AverageIntensity 22099 non-null float64 dtypes: float64(1), int64(2), object(1) memory usage: 690.7+ KB
In [74]:
hourlySteps.info()
<class 'pandas.core.frame.DataFrame'> RangeIndex: 22099 entries, 0 to 22098 Data columns (total 3 columns): # Column Non-Null Count Dtype --- ------ -------------- ----- 0 Id 22099 non-null int64 1 ActivityHour 22099 non-null object 2 StepTotal 22099 non-null int64 dtypes: int64(2), object(1) memory usage: 518.1+ KB
In [75]:
sleepDay.info()
<class 'pandas.core.frame.DataFrame'> RangeIndex: 413 entries, 0 to 412 Data columns (total 5 columns): # Column Non-Null Count Dtype --- ------ -------------- ----- 0 Id 413 non-null int64 1 SleepDay 413 non-null object 2 TotalSleepRecords 413 non-null int64 3 TotalMinutesAsleep 413 non-null int64 4 TotalTimeInBed 413 non-null int64 dtypes: int64(4), object(1) memory usage: 16.3+ KB
In [76]:
weightLogInfo.info()
<class 'pandas.core.frame.DataFrame'> RangeIndex: 67 entries, 0 to 66 Data columns (total 8 columns): # Column Non-Null Count Dtype --- ------ -------------- ----- 0 Id 67 non-null int64 1 Date 67 non-null object 2 WeightKg 67 non-null float64 3 WeightPounds 67 non-null float64 4 Fat 2 non-null float64 5 BMI 67 non-null float64 6 IsManualReport 67 non-null bool 7 LogId 67 non-null int64 dtypes: bool(1), float64(4), int64(2), object(1) memory usage: 3.9+ KB
In [77]:
# Dropping 'Fat' column in weightLogInfo since it is missing most of its values
weightLogInfo.drop('Fat', axis=1, inplace=True)
Data type conversion ¶
In [78]:
# Transforming the generic object columns containing dates into datetime64
dailyActivity['ActivityDate'] = dailyActivity['ActivityDate'].astype('datetime64[ns]')
hourlyIntensities['ActivityHour'] = hourlyIntensities['ActivityHour'].astype('datetime64[ns]')
hourlySteps['ActivityHour'] = hourlySteps['ActivityHour'].astype('datetime64[ns]')
sleepDay['SleepDay'] = sleepDay['SleepDay'].astype('datetime64[ns]')
weightLogInfo['Date'] = weightLogInfo['Date'].astype('datetime64[ns]')
Duplicated values ¶
In [79]:
# Dropping duplicates
dailyActivity = dailyActivity.drop_duplicates()
hourlyIntensities = hourlyIntensities.drop_duplicates()
hourlySteps = hourlySteps.drop_duplicates()
sleepDay = sleepDay.drop_duplicates()
weightLogInfo = weightLogInfo.drop_duplicates()
Analysis ¶
In [80]:
# There are 33 unique IDs
len(dailyActivity['Id'].unique())
Out[80]:
33
In [81]:
# The data has been taken in the time window between 4/12/2016 and 5/12/2016 (m/d/y)
min(dailyActivity['ActivityDate'])
Out[81]:
Timestamp('2016-04-12 00:00:00')
In [82]:
max(dailyActivity['ActivityDate'])
Out[82]:
Timestamp('2016-05-12 00:00:00')
Daily activity tracking ¶
In [83]:
# Out of the 33 users, 27 have kept the device on for at least
# 28 days out of 31 (more than 90% of the whole time window).
daily_tracking = {}
for user in dailyActivity['Id'].unique():
daily_tracking[user] = len(dailyActivity.query('Id == @user')['Id'])
fig1, ax1 = plt.subplots()
ax1.hist(daily_tracking.values(),
edgecolor = "black",
bins = [1,4,7,10,13,16,19,22,25,28,31])
plt.xticks([1,4,7,10,13,16,19,22,25,28,31]);
Daily activity intensity ¶
In [84]:
# Out of the 33 users, 18 have kept on average their device on them
# for more than 1296 minutes (more than 90% of the day duration)
daily_activity_intensity = {}
daily_activity_intensity_groupby = (pd.DataFrame(dailyActivity, columns=['Id', 'VeryActiveMinutes', 'FairlyActiveMinutes', 'LightlyActiveMinutes', 'SedentaryMinutes'])
.groupby(['Id'])
.mean()
.reset_index())
for intensity in ['VeryActiveMinutes', 'FairlyActiveMinutes','LightlyActiveMinutes', 'SedentaryMinutes']:
daily_activity_intensity[intensity] = list(daily_activity_intensity_groupby[intensity])
fig2, ax2 = plt.subplots()
ax2.bar([x for x in range(1,34)],
daily_activity_intensity['VeryActiveMinutes'],
label='VeryActiveMinutes',
edgecolor="black")
ax2.bar([x for x in range(1,34)],
daily_activity_intensity['FairlyActiveMinutes'],
bottom=daily_activity_intensity['VeryActiveMinutes'],
label='FairlyActiveMinutes',
edgecolor="black")
ax2.bar([x for x in range(1,34)],
daily_activity_intensity['LightlyActiveMinutes'],
bottom=[daily_activity_intensity['VeryActiveMinutes'][i] + daily_activity_intensity['FairlyActiveMinutes'][i] for i in range(len(daily_activity_intensity['VeryActiveMinutes']))],
label='LightlyActiveMinutes',
edgecolor="black")
ax2.bar([x for x in range(1,34)],
daily_activity_intensity['SedentaryMinutes'],
bottom=[daily_activity_intensity['VeryActiveMinutes'][i] + daily_activity_intensity['FairlyActiveMinutes'][i] + daily_activity_intensity['LightlyActiveMinutes'][i] for i in range(len(daily_activity_intensity['VeryActiveMinutes']))],
label='SedentaryMinutes',
edgecolor="black")
plt.axhline(1296,
color="black",
linestyle="dashed",
linewidth=1)
plt.xticks([x for x in range(1,34)], rotation=70)
ax2.legend(bbox_to_anchor=(1,1));
Sleep activity tracking ¶
In [85]:
# Out of 33 users, 24 have kept their device during sleep at least
# once, only 7 have kept it for at least 28 days. This confirms the
# prior hypotesis.
sleep_tracking = {}
for user in sleepDay['Id'].unique():
sleep_tracking[user] = len(sleepDay.query('Id == @user')['Id'])
fig3, ax3 = plt.subplots()
ax3.hist(sleep_tracking.values(),
edgecolor = "black",
bins = [1,4,7,10,13,16,19,22,25,28,31])
plt.xticks([1,4,7,10,13,16,19,22,25,28,31]);
Weight information tracking ¶
In [86]:
# Out of the 33 users, 8 have tracked their weight at least once,
# only 1 has kept track of their weight for at least 28 days.
weight_tracking = {}
for user in weightLogInfo['Id'].unique():
weight_tracking[user] = len(weightLogInfo.query('Id == @user')['Id'])
fig4, ax4 = plt.subplots()
ax4.hist(weight_tracking.values(),
edgecolor = "black",
bins = [1,4,7,10,13,16,19,22,25,28,31])
plt.xticks([1,4,7,10,13,16,19,22,25,28,31]);
Minutes of average total daily activity per user ¶
In [87]:
# Out of 33 users, 24 (more than 72% of the whole group)
# have more than 200 minutes of average total daily activity
total_activity = {}
for user in daily_activity_intensity_groupby['Id'].unique():
total_activity[user] = sum([daily_activity_intensity_groupby.query('Id == @user')['VeryActiveMinutes'].values[0],
daily_activity_intensity_groupby.query('Id == @user')['FairlyActiveMinutes'].values[0],
daily_activity_intensity_groupby.query('Id == @user')['LightlyActiveMinutes'].values[0]]
)
fig5, ax5 = plt.subplots()
ax5.hist(total_activity.values(),
edgecolor = "black",
bins = [0, 50, 100, 150, 200, 250, 300, 350]);
Hourly steps distribution ¶
In [88]:
# Users seem to be more incline to have more steps past
# the usual working hours during weekdays and before the
# usual working hours during weekends.
hourlySteps.insert(3, 'Hour', [pd.Timestamp(hourlySteps['ActivityHour'][i]).hour for i in hourlySteps.index])
hourlySteps.insert(4, 'Weekday', [pd.Timestamp(hourlySteps['ActivityHour'][i]).day_name() for i in hourlySteps.index])
hourly_weekdays = hourlySteps[(hourlySteps['Weekday'] == 'Monday') | (hourlySteps['Weekday'] == 'Tuesday') |
(hourlySteps['Weekday'] == 'Wednesday') | (hourlySteps['Weekday'] == 'Thursday') |
(hourlySteps['Weekday'] == 'Friday')]
hourly_weekend = hourlySteps[(hourlySteps['Weekday'] == 'Saturday') | (hourlySteps['Weekday'] == 'Sunday')]
hourly_weekdays_groupby = (pd.DataFrame(hourly_weekdays,
columns=['StepTotal','Hour'])
.groupby(['Hour'])
.mean())
hourly_weekend_groupby = (pd.DataFrame(hourly_weekend,
columns=['StepTotal','Hour'])
.groupby(['Hour'])
.mean())
fig7, ax7 = plt.subplots()
ax7.plot([i for i in hourly_weekdays_groupby.index],
list(hourly_weekdays_groupby['StepTotal']),
label='weekdays')
ax7.plot([i for i in hourly_weekend_groupby.index],
list(hourly_weekend_groupby['StepTotal']),
label='weekend')
plt.xticks([0, 4, 8, 12, 16, 20, 24])
plt.legend();
Activity intensity distribution ¶
In [89]:
# Users are mostly engaging light activities.
fig8, ax8 = plt.subplots()
ax8.bar([x for x in range(1,34)],
daily_activity_intensity['VeryActiveMinutes'],
label='VeryActiveMinutes',
edgecolor="black")
ax8.bar([x for x in range(1,34)],
daily_activity_intensity['FairlyActiveMinutes'],
bottom=daily_activity_intensity['VeryActiveMinutes'],
label='FairlyActiveMinutes',
edgecolor="black")
ax8.bar([x for x in range(1,34)],
daily_activity_intensity['LightlyActiveMinutes'],
bottom=[daily_activity_intensity['VeryActiveMinutes'][i] + daily_activity_intensity['FairlyActiveMinutes'][i] for i in range(len(daily_activity_intensity['VeryActiveMinutes']))],
label='LightlyActiveMinutes',
edgecolor="black")
plt.legend(bbox_to_anchor=(1,1))
plt.xticks([x for x in range(1,34)], rotation=70);
Conclusions ¶
Most of the users have kept their device on during the 30 days. This could mean that users are actually interested in the device keeping track of their activities and calories or at least interested in parameters that could be tracked automatically.
The latter hypotesis is also supported by the information gained from the analysis of the sleepDay dataset, where it is clear that not everyone is interested or want their sleep activities being tracked, and the weightLogInfo dataset, where very few people tracked their weight at least once.
What Bellabeat could do is manufacture a device that tracks daily activity and calories while also elaborate further through the use of surveys the interest of people in tracking other parameters such as sleeping activity and tracking weight to understand the reasons behind people not using such features as much.
The information found through the analysis of the average minutes of activities during the week may point towards the users of the device being mainly working people since the activity minutes seem to increase after the usual working hours while during the weekend the trend seem to be the opposite: more minutes during working hours and less after. Together with the information of the amount of activity done on average varying a lot from person to person, this could lead to a scenario where users using this device fall under different profiles and therefore be interested in custom plans that would optimize the time they have at their disposal.