Skip to content

03 – Exploratory Data Analysis

📘 RIDE User Manual – Panel 3: Exploratory Data Analysis (EDA)

📊 Purpose of the Panel

The EDA Panel helps users visualize patterns, distributions, and relationships in their data using a suite of 15+ interactive plot types. These visuals make it easier to understand underlying structures before performing modeling or transformation.

Recommended Reading

🧭 User Workflow

  1. Upload Dataset & Select Preprocessed Version
    User selects from:

    • Initial DataFrame
    • After Missing Value Imputation
    • After Feature Scaling
    • After Feature Encoding

  2. Header & Overview
    GIF and a markdown introduction help the user understand what the EDA panel does.

  3. Visual Grid
    Users are presented with a categorized set of plotting options:

    • Basic Plots
    • Advanced Plots
    • Specialized Plots
    • Geospatial Visualizations

  4. Click to Generate
    On clicking a plot type:

    • The corresponding visualization function is invoked.
    • The selected plot appears with a title and a back button.

💻 Features Breakdown

Feature Description
Data Source Selector Allows switching between raw and processed versions of the dataset.
Plot Grid Shows image previews and labels of 15+ supported plots.
One-click Plotting Clicking a plot button immediately renders the visual.
Plot Functions Powered by modular functions like plot_histogram, plot_boxplot, etc.
Back Navigation Users can return to the grid easily via a back button.

📊 Visualization Types Supported

🔹 Basic Plots

Plot Type Purpose / Use Case
Box Plot Shows the spread and skewness of data via quartiles and outliers. ✅ Best for comparing feature distributions, spotting outliers.
Histogram Displays frequency distribution of numeric values. ✅ Helps detect skewness, modality, and data concentration.
Scatter Plot Visualizes relationships between two continuous variables. ✅ Useful for detecting correlations and clusters.
Bar Chart Compares quantities of categorical variables. ✅ Ideal for viewing counts or aggregations across categories.
Pie Chart Shows proportional breakdown of categorical variables. ✅ Good for visualizing share/percentage composition.
Line Plot Displays trends across time or ordered observations. ✅ Best for time-series or sequential data analysis.

🔸 Advanced Plots

Plot Type Purpose / Use Case
2D Hist Contour Combines density-based histogram with contour lines. ✅ Useful when you want to detect dense regions in bivariate data.
Contour Plot Displays 3D surface on a 2D plane using contours. ✅ Ideal for understanding gradient/response surfaces.
Violin Plot Combines box plot and kernel density plot. ✅ Great for understanding the distribution shape and symmetry.
3D Scatter Visualizes 3D relationships between 3 continuous variables. ✅ Useful for dimensional analysis and observing clusters.
3D Line Connects points in 3D space. ✅ Helpful for time-based sequences across three dimensions.

🎯 Specialized Plots

Plot Type Purpose / Use Case
Polar Scatter Plots data on a circular axis (angle and radius). ✅ Best for directional or cyclical data (e.g., wind direction, time).
Polar Bar Like a bar chart in a circular layout. ✅ Useful for periodic or circular metrics (e.g., sales across 12 months).

🌍 Geospatial Visualizations

Plot Type Purpose / Use Case
Scatter Map Plots points on a geographical map. ✅ Excellent for visualizing locations (e.g., store distribution, incident reports).
Choropleth Map Colors geographic areas based on a variable. ✅ Great for showing demographic density, income levels, or COVID spread.
Bubble Map Uses bubbles to show values over geographical coordinates. ✅ Perfect for showing magnitude across locations (e.g., sales, population).

🤖 AI Integration

While this panel does not directly use GPT, it is fully compatible with previous and subsequent AI-assisted panels where generated insights complement these visualizations.