The objective of this project is to analyze data collected from smartwatches to gain insights into users' health and activity patterns. By examining various metrics such as steps taken, distance traveled, and energy burned, we aim to understand user behavior and improve fitness tracking.
The dataset I am using is from Kaggle. It was collected from 30 female users of the Fitbit smartwatch.
The analysis involves several steps using Python libraries such as Pandas, NumPy, Matplotlib, and Plotly. The process includes:
-
Data Cleaning: Missing values and outliers are identified and handled to ensure the integrity of the analysis.
-
Exploratory Data Analysis (EDA): Various visualizations are created to explore the data. This includes:
- Histograms to understand the distribution of steps taken.
- Box plots to identify outliers in the distance traveled.
- Time series plots to analyze activity trends over time.
-
Statistical Analysis: Correlation analysis is performed to identify relationships between different metrics, such as the correlation between total steps and energy burned.
-
Insights Generation: Based on the analysis, insights are drawn regarding user activity patterns, such as peak activity times and average daily steps.
-
Visualization: Interactive visualizations using Plotly are created to present the findings in an engaging manner.
The analysis reveals significant insights into user activity patterns, including average daily steps and the impact of sedentary behavior on overall health. For a detailed view of the analysis and visualizations, you can access the Jupyter Notebook here.