How vedheshvlv/sales-data-analysis-prediction works
This is a single-script retail sales forecasting project. The documented workflow loads and cleans the sales CSV, explores the data, encodes selected features, splits the data, trains four regressors, evaluates them, and visualizes comparisons and predictions. The script excerpt supports these internal stages, though they are not separate modules. The README also documents business interpretation and feature-importance analysis; source-level evidence for the latter is present in the script. No external services or persistent output stores are shown.
Main entry points: analysis.py Load & Clean · sales.csv Sales Dataset
Components 12
Data Preparation
- Sales Dataset
sales.csv - Load & Clean
analysis.py
Exploration & Features
- Sales EDA
analysis.py - Feature Encoding
analysis.py - Train-Test Split
analysis.py
Forecast Models
- Four Regressors
analysis.py
Evaluation & Insights
- Metric Evaluation
analysis.py - Model Comparison
analysis.py - Feature Importance
analysis.py - Prediction Plot
analysis.py
Other components
- Project User
- Script Runner
Connections 11
- Project User → Script Runner: runs script
README.md - Script Runner → Load & Clean: executes
README.md - Load & Clean → Sales Dataset: reads
analysis.py - Load & Clean → Sales EDA: provides cleaned data
analysis.py - Load & Clean → Feature Encoding: provides cleaned data
analysis.py - Feature Encoding → Train-Test Split: passes encoded data
analysis.py - Train-Test Split → Four Regressors: supplies train-test sets
analysis.py - Four Regressors → Metric Evaluation: provides predictions
analysis.py - Metric Evaluation → Model Comparison: supplies MAE RMSE
analysis.py - Four Regressors → Feature Importance: supplies forest model
analysis.py - Four Regressors → Prediction Plot: supplies predictions
analysis.py