Embedded Files
Fitness Tracking Analysis Project
Fitness Tracking Analysis Project
Objective
Objective
Examine available public data on fitness tracker devices to determinehow those devices are used. Identify from that data a marketing strategy for a fitness tracker device called Leaf by Bellabeat.
Data
Data
All Data used from FitBit Fitness Tracker Data by Amazon Mechanical Turk collected via survey between 3/12/2016- 5/12/2016. Public Domain, Preprocessed Data accessed via Kaggle.
FitBit Dataset is organized into 18 csv files. I used the dailyActivity_merged csv because it best applies to the Leaf fitness tracker device.
Portfolio_Project_R_ScriptData Processing
Data Processing
Imported Data into RStudio
I renamed every column using clean_names() and colnames(), ensured that there were no missing values using is.na(), created a new column using mutate(), and examined data using head() and skim_without_charts().
Please click the document to the left to view my R code used to clean and analyze my data.
Data Analysis
Data Analysis
I created a new dataframe containing the columns I would use in my analysis. These were ID, Total Steps, Tracker Distance, Calories, Very Active Minutes, Fairly Active Minutes, Lightly Active Minutes, Sedentary Minutes, and Total Active Minutes.
I used ggplot2 to create four visualizations that demonstrate the trends in usage of fitness tracker devices.
Steps and Calories
Steps and Calories
General trends of total steps influencing calories burned
As Total Steps increase, total calories increase
Positive Correlation
Distance and Steps
Distance and Steps
Trend of greater distance tracked occurring with more total steps
As distance tracked increases, total steps increase
Positive Correlation
Activity and Calories
Activity and Calories
This visualization shows the impact of active minutes on calories
Total of fairly, lightly, and active minutes added together to make total active minutes per person
Total active minutes shown in black dots in the scatterplot
Each subdivision of fairly, lightly, and very active minutes given their own line
Fairly Active Minutes has a slight increase then a decrease in calories burned while active minutes increase
Lightly Active Minutes shows a steady increase of calories burned with greater active minutes
Very Active Minutes shows a steep increase of calories burned while active minutes increase
Calories have a positive correlation with active minutes in the very active, lightly active, and total active minutes categories
Lightly Active Minutes is closest to the trend of total active minutes
Sedentary and Active Minutes
Sedentary and Active Minutes
Demonstrates relationship between active minutes and sedentary minutes
Fairly active minutes has similar active minutes among sedentary minute levels
Very active minutes has slightly fewer active minutes with higher sedentary minutes
Lightly active minutes has highest active minutes for its group at mid levels of sedentary minutes with a decrease beginning at 1,000 sedentary minutes
Total active minutes most closely represents the lightly active minutes plot
Slight negative correlation, greater at lightly and total active minutes categories above 1,000 sedentary minutes
Conclusions
Conclusions
Greater Active Minutes positively correlated with calories burned
Lightly Active Minutes most closely represents Total Active Minutes when looking at active minutes plotted with calories and sedentary minutes.
The Activity levels of Lightly Active Minutes are closest to the Total Activity Minutes.
Groups who participate in Lightly Active Minutes are closest to the total groups who use fitness tracking devices.
Marketing strategy should focus upon those who exercise with lightly active minutes. These are people working to increase calories burned and may want to lose weight. A general population who wants to improve their health by using a fitness tracker and participating in light exercise.
Page updated
Google Sites
Report abuse