Case Study For Smart Devices
Rashmi Shetty
2025-04-30
INTRODUCTION:
Bellabeat, a high-tech manufacturer of health-focused products for women. Collecting data on activity, sleep, stress, and reproductive health has allowed Bellabeat to empower women with knowledge about their own health and habits. Since it was founded in 2013, Bellabeat has grown rapidly and quickly positioned itself as a tech-driven wellness company for women. Bellabeat is a successful small company, but they have the potential to become a larger player in the global smart device market. Urška Sršen, co-founder and Chief Creative Officer of Bellabeat, believes that analyzing smart device fitness data could help unlock new growth opportunities for the company. So she wants us to analyze smart device usage data in order to gain insight into how consumers use non-Bellabeat smart devices. She then wants us to select one Bellabeat product to apply these insights to in our presentation.
Reports & recommendations to be submitted to:
- Urška Sršen: Bellabeat’s co-founder and Chief Creative Officer
- Sando Mur: Mathematician and Bellabeat’s co-founder; key member of the (Bellabeat Executive team)
- Bellabeat marketing analytics team: A team of data analysts responsible for collecting, analyzing, and reporting data that helps guide Bellabeat’s marketing strategy.
Business Task:
- Focus on one of the Bellabeat’s products- (Here I have chosen
Leaf: Bellabeat’s classic wellness tracker can be worn
as a bracelet, necklace, or clip. The Leaf tracker connects to the
Bellabeat app to track activity, sleep, and stress.)
- Focus on one of the Bellabeat’s products- (Here I have chosen
Leaf: Bellabeat’s classic wellness tracker can be worn
as a bracelet, necklace, or clip. The Leaf tracker connects to the
Bellabeat app to track activity, sleep, and stress.)
- 2.Analyze Fitbit smart device data to gain insight into how
consumers are using their smart devices-
- Define some trends in smart device usage?
- How could these trends apply to Bellabeat customers?
- How could these trends help influence Bellabeat marketing strategy?
- 3.Provide insights & high- level recommendations that help guide marketing strategy for the company, based on the analysis.
The Datasets reffered to for the analysis is https://www.kaggle.com/datasets/arashnic/fitbit
The datasets included here are:
- Daily activity dataset
- Sleep dataset
- Heart rate dataset
- Intensity/ minute dataset
- Steps/ minute dataset
- Calories/ minute dataset
- METs/ minute dataset
The datasets also include weight change & B.M.I tracking for very less customer Ids, so is omitted from the reports. Also all the data is of the year 2016, so its credibility for the current year can’t be stated. But Kaggle’s usabilty index of 8.75 is avalid proof of genuinness of the data, so we proceed with our data analysis.
So we prepare data for analysis by downloading the required packages
Preparing Data
Reading the dataset
daily <- fread("D:/med tour Easy files/useffu data/Case study files/Fitabase Data 3.12.16-4.11.16/dailyActivity_merged.csv")checking the dataset for null values
summary(daily)## Id ActivityDate TotalSteps TotalDistance
## Min. :1503960366 Length:457 Min. : 0 Min. : 0.000
## 1st Qu.:2347167796 Class :character 1st Qu.: 1988 1st Qu.: 1.410
## Median :4057192912 Mode :character Median : 5986 Median : 4.090
## Mean :4628594643 Mean : 6547 Mean : 4.664
## 3rd Qu.:6391747486 3rd Qu.:10198 3rd Qu.: 7.160
## Max. :8877689391 Max. :28497 Max. :27.530
## TrackerDistance LoggedActivitiesDistance VeryActiveDistance
## Min. : 0.00 Min. :0.0000 Min. : 0.000
## 1st Qu.: 1.28 1st Qu.:0.0000 1st Qu.: 0.000
## Median : 4.09 Median :0.0000 Median : 0.000
## Mean : 4.61 Mean :0.1794 Mean : 1.181
## 3rd Qu.: 7.11 3rd Qu.:0.0000 3rd Qu.: 1.310
## Max. :27.53 Max. :6.7271 Max. :21.920
## ModeratelyActiveDistance LightActiveDistance SedentaryActiveDistance
## Min. :0.0000 Min. : 0.00 Min. :0.000000
## 1st Qu.:0.0000 1st Qu.: 0.87 1st Qu.:0.000000
## Median :0.0200 Median : 2.93 Median :0.000000
## Mean :0.4786 Mean : 2.89 Mean :0.001904
## 3rd Qu.:0.6700 3rd Qu.: 4.46 3rd Qu.:0.000000
## Max. :6.4000 Max. :12.51 Max. :0.100000
## VeryActiveMinutes FairlyActiveMinutes LightlyActiveMinutes SedentaryMinutes
## Min. : 0.00 Min. : 0.00 Min. : 0.0 Min. : 32.0
## 1st Qu.: 0.00 1st Qu.: 0.00 1st Qu.: 64.0 1st Qu.: 728.0
## Median : 0.00 Median : 1.00 Median :181.0 Median :1057.0
## Mean : 16.62 Mean : 13.07 Mean :170.1 Mean : 995.3
## 3rd Qu.: 25.00 3rd Qu.: 16.00 3rd Qu.:257.0 3rd Qu.:1285.0
## Max. :202.00 Max. :660.00 Max. :720.0 Max. :1440.0
## Calories
## Min. : 0
## 1st Qu.:1776
## Median :2062
## Mean :2189
## 3rd Qu.:2667
## Max. :4562No null values in the dataset
Sorting the dataset & checking for duplicate values
sorted_df <- daily[order(daily), ]
head(sorted_df)## Id ActivityDate TotalSteps TotalDistance TrackerDistance
## <i64> <char> <int> <num> <num>
## 1: 1503960366 3/25/2016 11004 7.11 7.11
## 2: 1503960366 3/26/2016 17609 11.55 11.55
## 3: 1503960366 3/27/2016 12736 8.53 8.53
## 4: 1503960366 3/28/2016 13231 8.93 8.93
## 5: 1503960366 3/29/2016 12041 7.85 7.85
## 6: 1503960366 3/30/2016 10970 7.16 7.16
## LoggedActivitiesDistance VeryActiveDistance ModeratelyActiveDistance
## <num> <num> <num>
## 1: 0 2.57 0.46
## 2: 0 6.92 0.73
## 3: 0 4.66 0.16
## 4: 0 3.19 0.79
## 5: 0 2.16 1.09
## 6: 0 2.36 0.51
## LightActiveDistance SedentaryActiveDistance VeryActiveMinutes
## <num> <num> <int>
## 1: 4.07 0 33
## 2: 3.91 0 89
## 3: 3.71 0 56
## 4: 4.95 0 39
## 5: 4.61 0 28
## 6: 4.29 0 30
## FairlyActiveMinutes LightlyActiveMinutes SedentaryMinutes Calories
## <int> <int> <int> <int>
## 1: 12 205 804 1819
## 2: 17 274 588 2154
## 3: 5 268 605 1944
## 4: 20 224 1080 1932
## 5: 28 243 763 1886
## 6: 13 223 1174 1820No duplicate values
End of data cleaning(Process phase)
Analyzing data
distinct_Id <- sorted_df %>%
distinct(Id)
distinct_Id## Id
## <i64>
## 1: 1503960366
## 2: 1624580081
## 3: 1644430081
## 4: 1844505072
## 5: 1927972279
## 6: 2022484408
## 7: 2026352035
## 8: 2320127002
## 9: 2347167796
## 10: 2873212765
## 11: 2891001357
## 12: 3372868164
## 13: 3977333714
## 14: 4020332650
## 15: 4057192912
## 16: 4319703577
## 17: 4388161847
## 18: 4445114986
## 19: 4558609924
## 20: 4702921684
## 21: 5553957443
## 22: 5577150313
## 23: 6117666160
## 24: 6290855005
## 25: 6391747486
## 26: 6775888955
## 27: 6962181067
## 28: 7007744171
## 29: 7086361926
## 30: 8053475328
## 31: 8253242879
## 32: 8378563200
## 33: 8583815059
## 34: 8792009665
## 35: 8877689391
## IdData contains 35 unique IDs
distinct_ad <- sorted_df %>%
distinct(ActivityDate)
distinct_ad## ActivityDate
## <char>
## 1: 3/25/2016
## 2: 3/26/2016
## 3: 3/27/2016
## 4: 3/28/2016
## 5: 3/29/2016
## 6: 3/30/2016
## 7: 3/31/2016
## 8: 4/1/2016
## 9: 4/10/2016
## 10: 4/11/2016
## 11: 4/12/2016
## 12: 4/2/2016
## 13: 4/3/2016
## 14: 4/4/2016
## 15: 4/5/2016
## 16: 4/6/2016
## 17: 4/7/2016
## 18: 4/8/2016
## 19: 4/9/2016
## 20: 3/12/2016
## 21: 3/13/2016
## 22: 3/14/2016
## 23: 3/15/2016
## 24: 3/16/2016
## 25: 3/17/2016
## 26: 3/18/2016
## 27: 3/19/2016
## 28: 3/20/2016
## 29: 3/21/2016
## 30: 3/22/2016
## 31: 3/23/2016
## 32: 3/24/2016
## ActivityDateData contains 32 unique activity dates
Now lets analyze the parameters used for tracking fitness data by choosing a single ID- 2347167796
filtered_df <- sorted_df[sorted_df$Id == 2347167796, ]
print(filtered_df)## Id ActivityDate TotalSteps TotalDistance TrackerDistance
## <i64> <char> <int> <num> <num>
## 1: 2347167796 3/29/2016 10272 6.79 6.79
## 2: 2347167796 3/30/2016 10533 7.10 7.10
## 3: 2347167796 3/31/2016 6760 4.47 4.47
## 4: 2347167796 4/1/2016 8328 5.51 5.51
## 5: 2347167796 4/10/2016 10078 6.83 6.83
## 6: 2347167796 4/11/2016 10001 6.61 6.61
## 7: 2347167796 4/12/2016 0 0.00 0.00
## 8: 2347167796 4/2/2016 15459 10.22 10.22
## 9: 2347167796 4/3/2016 7485 4.95 4.95
## 10: 2347167796 4/4/2016 10254 6.80 6.80
## 11: 2347167796 4/5/2016 10114 6.82 6.82
## 12: 2347167796 4/6/2016 11107 7.34 7.34
## 13: 2347167796 4/7/2016 10320 6.85 6.85
## 14: 2347167796 4/8/2016 10209 6.75 6.75
## 15: 2347167796 4/9/2016 16081 10.63 10.63
## LoggedActivitiesDistance VeryActiveDistance ModeratelyActiveDistance
## <num> <num> <num>
## 1: 0 0.16 3.12
## 2: 0 1.77 2.06
## 3: 0 0.00 0.00
## 4: 0 0.00 2.00
## 5: 0 1.02 0.12
## 6: 0 0.33 2.93
## 7: 0 0.00 0.00
## 8: 0 3.59 0.81
## 9: 0 0.00 0.00
## 10: 0 1.42 1.23
## 11: 0 1.64 0.48
## 12: 0 0.90 2.76
## 13: 0 0.68 1.23
## 14: 0 0.16 0.35
## 15: 0 1.25 1.82
## LightActiveDistance SedentaryActiveDistance VeryActiveMinutes
## <num> <num> <int>
## 1: 3.50 0 2
## 2: 3.27 0 21
## 3: 4.47 0 0
## 4: 3.50 0 0
## 5: 5.69 0 12
## 6: 3.36 0 5
## 7: 0.00 0 0
## 8: 5.82 0 51
## 9: 4.95 0 0
## 10: 4.16 0 21
## 11: 4.68 0 18
## 12: 3.68 0 14
## 13: 4.94 0 15
## 14: 6.23 0 2
## 15: 7.56 0 16
## FairlyActiveMinutes LightlyActiveMinutes SedentaryMinutes Calories
## <int> <int> <int> <int>
## 1: 58 208 700 2041
## 2: 35 255 615 2187
## 3: 0 250 613 1929
## 4: 33 212 804 1935
## 5: 3 303 463 2164
## 6: 52 196 788 2009
## 7: 0 0 425 399
## 8: 16 327 583 2438
## 9: 0 324 491 2035
## 10: 25 231 638 2099
## 11: 9 269 696 2096
## 12: 46 196 759 2058
## 13: 32 315 1002 2338
## 14: 6 316 711 2104
## 15: 32 401 970 2488Take the mean of all variables
1>
actmin_df <- filtered_df[, lapply(.SD, mean), .SDcols = c("VeryActiveMinutes","FairlyActiveMinutes",
"LightlyActiveMinutes", "SedentaryMinutes")]
actmin_df ## VeryActiveMinutes FairlyActiveMinutes LightlyActiveMinutes SedentaryMinutes
## <num> <num> <num> <num>
## 1: 11.8 23.13333 253.5333 683.8667create a ggplot from the data to determine active minutes in a day
active_minutes_in_a_day <- melt(actmin_df,
measure.vars = c("VeryActiveMinutes", "FairlyActiveMinutes",
"LightlyActiveMinutes", "SedentaryMinutes"),
variable.name = "ActivityLevel",
value.name = "AverageMinutes")ggplot(active_minutes_in_a_day, aes(x = ActivityLevel, y = AverageMinutes, fill = ActivityLevel)) +
geom_bar(stat = "identity") +
geom_text(aes(label = round(AverageMinutes, 1)), vjust = -0.5) +
labs(title = "Average Daily Activity Minutes by Type", subtitle = "Activity level in a day",
x = "Activity Level",
y = "Average Minutes") +
theme_minimal() +
theme(legend.position = "middle") 
The chart shows that most of the time is spent Sedentarily, No much active minutes
2>
create a ggplot from the data to determine active distance in a day
actdis_df <- filtered_df[, lapply(.SD, mean), .SDcols = c("VeryActiveDistance", "ModeratelyActiveDistance", "LightActiveDistance", "SedentaryActiveDistance")]
actdis_df ## VeryActiveDistance ModeratelyActiveDistance LightActiveDistance
## <num> <num> <num>
## 1: 0.8613333 1.260667 4.387333
## SedentaryActiveDistance
## <num>
## 1: 0active_distance_in_a_day <- melt(actdis_df, measure.vars
= c("VeryActiveDistance", "ModeratelyActiveDistance", "LightActiveDistance", "SedentaryActiveDistance"),
variable.name = "ActivityLevel",
value.name = "AverageDistance")active_distance_in_a_day[, Percentage := AverageDistance / sum(AverageDistance) * 100]ggplot(active_distance_in_a_day, aes(x = "Activity", y = Percentage, fill = ActivityLevel)) +
geom_bar(stat = "identity", width = 0.5) +
geom_text(aes(label = paste0(round(Percentage, 1), "%")),
position = position_stack(vjust = 0.5), color = "white", size = 4) +
labs(title = "Proportion of Daily Activity by Type",
x = NULL,
y = "Percentage") +
theme_minimal()+
theme(axis.text.x = element_blank(),
axis.ticks.x = element_blank(),
legend.title = element_blank())
The chart shows more Light Active Distance than active distance
Also here we attach a pie chart depicting the clear picture
Pie chart plotting Activity level
Now lets understand the co-relation between total steps & calories:
ggplot(filtered_df)+ geom_line(mapping= aes( x= Calories, y= TotalSteps, color ="calories"))+
labs(title= "Corelation of steps & Calories")+
theme_minimal()
##### The plot shows a positive correlation between Total steps &
calories burnt.
Now lets understand the co-relation between total Distance & calories:
ggplot(filtered_df)+ geom_line(mapping= aes( x= Calories, y= TotalDistance, color ="calories"))+
labs(title= "Corelation of Distance & Calories")+
theme_minimal()
##### The plot shows a positive correlation between Total steps &
calories burnt.
Reading Sleep value dataset:
sleep <- fread("D:/med tour Easy files/useffu data/Case study files/Fitabase Data 3.12.16-4.11.16/minuteSleep_merged.csv")
print(sleep)## Id date value log_id
## <i64> <char> <int> <i64>
## 1: 1503960366 3/13/2016 2:39 1 11114919637
## 2: 1503960366 3/13/2016 2:40 1 11114919637
## 3: 1503960366 3/13/2016 2:41 1 11114919637
## 4: 1503960366 3/13/2016 2:42 1 11114919637
## 5: 1503960366 3/13/2016 2:43 1 11114919637
## ---
## 198555: 8792009665 4/9/2016 18:38 1 11357751881
## 198556: 8792009665 4/9/2016 18:39 1 11357751881
## 198557: 8792009665 4/9/2016 18:40 1 11357751881
## 198558: 8792009665 4/9/2016 18:41 1 11357751881
## 198559: 8792009665 4/9/2016 18:42 1 11357751881sleep_filtered <- sleep[sleep$Id == 2347167796, ]
sleep_filtered## Id date value log_id
## <i64> <char> <int> <i64>
## 1: 2347167796 3/11/2016 23:43 3 11103744323
## 2: 2347167796 3/11/2016 23:44 3 11103744323
## 3: 2347167796 3/11/2016 23:45 2 11103744323
## 4: 2347167796 3/11/2016 23:46 1 11103744323
## 5: 2347167796 3/11/2016 23:47 1 11103744323
## ---
## 13480: 2347167796 4/11/2016 6:34 2 11363107879
## 13481: 2347167796 4/11/2016 6:35 3 11363107879
## 13482: 2347167796 4/11/2016 6:36 2 11363107879
## 13483: 2347167796 4/11/2016 6:37 2 11363107879
## 13484: 2347167796 4/11/2016 6:38 2 11363107879sleep_filtered$date <- as.POSIXct(sleep_filtered$date, format = "%m/%d/%Y %H:%M")Extract date in yyyy/mm/dd format
sleep_filtered$Date_only <- format(sleep_filtered$date, "%Y/%m/%d")Extract time in hh:mm:ss format
sleep_filtered$Time_only <- format(sleep_filtered$date, "%H:%M:%S")
sleep_filtered## Id date value log_id Date_only Time_only
## <i64> <POSc> <int> <i64> <char> <char>
## 1: 2347167796 2016-03-11 23:43:00 3 11103744323 2016/03/11 23:43:00
## 2: 2347167796 2016-03-11 23:44:00 3 11103744323 2016/03/11 23:44:00
## 3: 2347167796 2016-03-11 23:45:00 2 11103744323 2016/03/11 23:45:00
## 4: 2347167796 2016-03-11 23:46:00 1 11103744323 2016/03/11 23:46:00
## 5: 2347167796 2016-03-11 23:47:00 1 11103744323 2016/03/11 23:47:00
## ---
## 13480: 2347167796 2016-04-11 06:34:00 2 11363107879 2016/04/11 06:34:00
## 13481: 2347167796 2016-04-11 06:35:00 3 11363107879 2016/04/11 06:35:00
## 13482: 2347167796 2016-04-11 06:36:00 2 11363107879 2016/04/11 06:36:00
## 13483: 2347167796 2016-04-11 06:37:00 2 11363107879 2016/04/11 06:37:00
## 13484: 2347167796 2016-04-11 06:38:00 2 11363107879 2016/04/11 06:38:00sleep_filtered <- as.data.table(sleep_filtered)sleep_filtered[, Date_only := as.Date(date)] Calculate mean Value per distinct Date_only:
daily_mean <- sleep_filtered[, .(Mean_Value = mean(value, na.rm = TRUE)),
by = Date_only]
daily_mean## Date_only Mean_Value
## <Date> <num>
## 1: 2016-03-11 1.043228
## 2: 2016-03-12 1.094188
## 3: 2016-03-13 1.082258
## 4: 2016-03-14 1.121795
## 5: 2016-03-15 1.079855
## 6: 2016-03-16 1.108384
## 7: 2016-03-17 1.090186
## 8: 2016-03-18 1.272517
## 9: 2016-03-19 1.100000
## 10: 2016-03-20 1.100855
## 11: 2016-03-21 1.088843
## 12: 2016-03-22 1.074570
## 13: 2016-03-23 1.131846
## 14: 2016-03-24 1.126273
## 15: 2016-03-25 1.000000
## 16: 2016-03-27 1.130802
## 17: 2016-03-28 1.133929
## 18: 2016-03-29 1.042373
## 19: 2016-03-30 1.126459
## 20: 2016-03-31 1.081456
## 21: 2016-04-01 1.094556
## 22: 2016-04-02 1.116832
## 23: 2016-04-03 1.083200
## 24: 2016-04-04 1.139048
## 25: 2016-04-05 1.176339
## 26: 2016-04-06 1.084211
## 27: 2016-04-07 1.083744
## 28: 2016-04-08 1.013333
## 29: 2016-04-09 1.099715
## 30: 2016-04-10 1.116844
## 31: 2016-04-11 1.130435
## Date_only Mean_Valueggplot(daily_mean)+ geom_line(mapping= aes( x= Date_only, y= Mean_Value), color ="Blue")+
labs(title= "Sleep value fluctuation through the days of tracking")+
theme_minimal()
Smoother the sleep plot, smoother is that indivisuals sleep. The above plot indicates improper sleeping patterns or no healthy sleep.
Time to read heart rate dataset
heart <- fread("D:/med tour Easy files/useffu data/Case study files/Fitabase Data 3.12.16-4.11.16/heartrate_seconds_merged.csv")
print(heart)## Id Time Value
## <i64> <char> <int>
## 1: 2022484408 4/1/2016 7:54 93
## 2: 2022484408 4/1/2016 7:54 91
## 3: 2022484408 4/1/2016 7:54 96
## 4: 2022484408 4/1/2016 7:54 98
## 5: 2022484408 4/1/2016 7:54 100
## ---
## 1048571: 8792009665 4/7/2016 14:20 80
## 1048572: 8792009665 4/7/2016 14:21 80
## 1048573: 8792009665 4/7/2016 14:21 79
## 1048574: 8792009665 4/7/2016 14:21 79
## 1048575: 8792009665 4/7/2016 14:21 80heart_filtered <- heart[heart$Id == 2347167796, ]
heart_filtered## Id Time Value
## <i64> <char> <int>
## 1: 2347167796 3/29/2016 0:00 69
## 2: 2347167796 3/29/2016 0:00 68
## 3: 2347167796 3/29/2016 0:00 69
## 4: 2347167796 3/29/2016 0:00 69
## 5: 2347167796 3/29/2016 0:00 69
## ---
## 120800: 2347167796 4/11/2016 21:45 81
## 120801: 2347167796 4/11/2016 21:45 83
## 120802: 2347167796 4/11/2016 21:45 84
## 120803: 2347167796 4/11/2016 21:45 86
## 120804: 2347167796 4/11/2016 21:46 86heart_filtered$Time <- as.POSIXct(heart_filtered$Time, format = "%m/%d/%Y %H:%M")Extract date in yyyy/mm/dd format
heart_filtered$Date_only <- format(heart_filtered$Time, "%Y/%m/%d")Extract time in hh:mm:ss format
heart_filtered$Time_only <- format(heart_filtered$Time, "%H:%M:%S")
heart_filtered## Id Time Value Date_only Time_only
## <i64> <POSc> <int> <char> <char>
## 1: 2347167796 2016-03-29 00:00:00 69 2016/03/29 00:00:00
## 2: 2347167796 2016-03-29 00:00:00 68 2016/03/29 00:00:00
## 3: 2347167796 2016-03-29 00:00:00 69 2016/03/29 00:00:00
## 4: 2347167796 2016-03-29 00:00:00 69 2016/03/29 00:00:00
## 5: 2347167796 2016-03-29 00:00:00 69 2016/03/29 00:00:00
## ---
## 120800: 2347167796 2016-04-11 21:45:00 81 2016/04/11 21:45:00
## 120801: 2347167796 2016-04-11 21:45:00 83 2016/04/11 21:45:00
## 120802: 2347167796 2016-04-11 21:45:00 84 2016/04/11 21:45:00
## 120803: 2347167796 2016-04-11 21:45:00 86 2016/04/11 21:45:00
## 120804: 2347167796 2016-04-11 21:46:00 86 2016/04/11 21:46:00heart_filtered <- as.data.table(heart_filtered)heart_filtered[, Date_only := as.Date(Time)] # if not already doneCalculate mean Value per distinct Date_only
Filter and group
daily_mean <- heart_filtered[, .(Mean_Value = mean(Value, na.rm = TRUE)),
by = Date_only]
daily_mean## Date_only Mean_Value
## <Date> <num>
## 1: 2016-03-28 64.79339
## 2: 2016-03-29 70.02429
## 3: 2016-03-30 73.92168
## 4: 2016-03-31 70.10231
## 5: 2016-04-01 71.88555
## 6: 2016-04-02 81.00350
## 7: 2016-04-03 74.40492
## 8: 2016-04-04 75.59180
## 9: 2016-04-05 76.25823
## 10: 2016-04-06 76.08930
## 11: 2016-04-07 83.38784
## 12: 2016-04-08 77.32405
## 13: 2016-04-09 83.99412
## 14: 2016-04-10 78.04957
## 15: 2016-04-11 75.49036Plot Heart rate fluctuation throughout the Day:
ggplot(heart_filtered)+ geom_line(mapping= aes( x= Value, y= Time_only), color = "Red")+
labs(title= "Heart rate fluctuation throughout the day")+
theme_minimal()
Since the readings taken are per second, we get to see a plot showing vibrant peaks & lows
Indicating that the individual is healthy that particular day
Its time to read another dataset, a merged dataset of 5 datasets
hourly intensity, hourly steps & hourly calories with heart rate & METs.
I’ve already merged them in SQL
calling data & inspecting it
HMISC <- fread("D:/med tour Easy files/useffu data/Case study files/Fitabase Data 3.12.16-4.11.16/newly joined data/heart_MISC_minutes_data.csv")
print(HMISC)## Id Activity_date Activity_time Value Intensity Steps METs
## <i64> <char> <char> <int> <int> <int> <int>
## 1: 2347167796 3/29/2016 00:55:00.000000 UTC 69 1 5 24
## 2: 2347167796 3/29/2016 00:55:00.000000 UTC 70 1 5 24
## 3: 2347167796 3/29/2016 00:55:00.000000 UTC 71 1 5 24
## 4: 2347167796 3/29/2016 00:55:00.000000 UTC 72 1 5 24
## 5: 2347167796 3/29/2016 00:55:00.000000 UTC 73 1 5 24
## ---
## 2400: 2347167796 3/29/2016 22:03:00.000000 UTC 62 1 5 24
## 2401: 2347167796 3/29/2016 22:03:00.000000 UTC 64 1 5 24
## 2402: 2347167796 3/29/2016 22:03:00.000000 UTC 65 1 5 24
## 2403: 2347167796 3/29/2016 22:03:00.000000 UTC 67 1 5 24
## 2404: 2347167796 3/29/2016 22:03:00.000000 UTC 70 1 5 24
## Calories
## <num>
## 1: 2.2584
## 2: 2.2584
## 3: 2.2584
## 4: 2.2584
## 5: 2.2584
## ---
## 2400: 2.2584
## 2401: 2.2584
## 2402: 2.2584
## 2403: 2.2584
## 2404: 2.2584summary(HMISC)## Id Activity_date Activity_time Value
## Min. :2347167796 Length:2404 Length:2404 Min. : 59.00
## 1st Qu.:2347167796 Class :character Class :character 1st Qu.: 76.00
## Median :2347167796 Mode :character Mode :character Median : 84.00
## Mean :2347167796 Mean : 84.48
## 3rd Qu.:2347167796 3rd Qu.: 92.00
## Max. :2347167796 Max. :135.00
## Intensity Steps METs Calories
## Min. :0.000 Min. : 0.0 Min. :12.00 Min. :1.129
## 1st Qu.:1.000 1st Qu.: 8.0 1st Qu.:24.00 1st Qu.:2.258
## Median :1.000 Median : 23.0 Median :30.00 Median :2.823
## Mean :1.152 Mean : 38.2 Mean :34.36 Mean :3.233
## 3rd Qu.:1.000 3rd Qu.: 73.0 3rd Qu.:46.00 3rd Qu.:4.329
## Max. :3.000 Max. :128.0 Max. :68.00 Max. :6.399There are no null values
so lets proceed analyzing the filtered data for ID- 2347167796
The dataset includes the per minute tracking details for Heart rate(value), Intensity, Steps, Calories & METs(Metabolic Equivalent of Task).
Next we proceed towards analyzing the relation between the above mentioned variables & how they give insight into the tracking trends of smart devices.
so lets proceed analyzing the filtered data for ID- 2347167796
Plot Intensity Vs. METs
ggplot(HMISC)+ geom_line(mapping= aes( x= Intensity, y= METs), color ="Green")+
labs(title= "Plot Intensity Vs. METs")+
theme_minimal()
Showing positive correlation
Chart Showing Positive correlation between Heart
rate Vs. Steps & METs
Analysis Summary:
- Trends The analysis shows that people use smart devices mostly for tracking steps, calories, intensity, heart rate, sleep & METs.
- We have filtered single Ids for some of our analysis to get a clear picture of the data
- Total steps & calories have a positive correlation.
- A healthy person’s heart rate chart shows clear peaks & lows.
- Activity level chart shows that people spend most of their time in inactivity(sedentary) rather than engaging in any form of physical activity.
- Higher METs values, healthier the individual.
Recommendations:
- Since Bellabeat’s customers are women, the design of the product ‘Leaf’ is absolutely perfect & trendy. Also the device can measure all the parameters that a fitbit device can measure. But one specific recommendation includes adding a alert system that notifies the customer for every 60 mins. of no activity(sedentary minutes). This feature will enable people to get alert & start some physical activity.
- Also alerting customers regarding their METs value will also help them keep track of their workout pattern, if they need to make any changes in their workout intensity to achieve desired fitness levels.
- Marketing Strategy Since the people using fitness tracker devices will be conscious about their health, Bellabeat marketing team can target gyms, sports academies, airport advertisements with repeated tv ads to target the potential customers. After all who would not want someone(device) who would remind us to be healthy & fit.