How hanfia-sattar/rossmore_sales_project works
This repository is a Rossmann store-sales analysis and forecasting project, with both a Streamlit dashboard and a command-line workflow. The command-line path loads sales data, explores and preprocesses it, runs forecasting, then generates business insights. A separate Streamlit app implements a store-series LSTM forecasting flow. The larger dashboard advertises preprocessing, model training, forecasts, and insights; its source is only partially excerpted, so unshown runtime links are omitted. The README is empty, and no external service integration is evident in the sampled material.
Main entry points: main.py CLI Workflow · forecasting_pipeline.py Forecast Pipeline · app.py LSTM App · data_preprocessing.py Data Preparation
Components 15
Analysis Workflow
- CLI Workflow
main.py - Exploratory Analysis
exploratory_analysis.py - Data Preparation
data_preprocessing.py - Business Insights
business_insights.py - Data Calculations
data_cal.py
User Interfaces
- Forecast Dashboard
dashboard.py - Dashboard Copy
dashboard_copy.py - LSTM App
app.py
Forecasting Models
- Forecast Pipeline
forecasting_pipeline.py - Hybrid Stacking
hybrid_stacking_model.py
Data and Artifacts
- Sales Data
Data/train.csv - Store Data
Data/store.csv - Forecast Results
sales_forecast_results.csv - Saved Model
rossmann_sales_forecast_random_forest_model.pkl
Other components
- Analyst
Connections 12
- Analyst → CLI Workflow: runs
main.py - CLI Workflow → Data Preparation: loads and prepares
main.py - CLI Workflow → Exploratory Analysis: runs analysis
main.py - CLI Workflow → Data Preparation: engineers features
main.py - CLI Workflow → Forecast Pipeline: runs forecasts
main.py - CLI Workflow → Business Insights: generates insights
main.py - Analyst → Forecast Dashboard: uses
- Analyst → Dashboard Copy: uses
- Analyst → LSTM App: uses
app.py - LSTM App → Sales Data: loads
app.py - Forecast Pipeline → Saved Model: saves model
forecasting_pipeline.py - Forecast Pipeline → Forecast Results: saves results
forecasting_pipeline.py