How brook1167/rossmann-pharmaceuticals works
Rossmann Pharmaceuticals is a sales-forecasting project, with exploratory analysis and model development alongside a Flask prediction endpoint. The documented user flow is an analyst submitting store and date features and receiving predicted sales. The endpoint’s preprocessing and model-loading imports are not present in the tree, so their implementations and any monitoring mechanism cannot be traced; links are limited to behavior visible in the supplied source and README.
Main entry points: app.py Flask application · modeling.py Model training · notebooks Exploratory analysis · data_processing.py Data processing
Components 15
Data analysis
- Exploratory analysis
notebooks - Data processing
scripts/data_processing.py - Sales visualizations
scripts/data_plotter.py
Forecast modeling
- Feature preparation
scripts/modeling.py - Model training
scripts/modeling.py - Model evaluation
scripts/modeling.py - LSTM forecasting
scripts/modeling.py
Prediction serving
- Flask application
app/app.py - Request preprocessing
- Loaded prediction model
app/app.py - Sales prediction
Model operations
- Prediction deployment
app - Performance monitoring
Other components
- Finance analyst
- Sales data
Connections 12
- Sales data → Exploratory analysis: analyzed in
- Sales data → Data processing: inspected and transformed
- Sales data → Sales visualizations: visualized
- Feature preparation → Model training: prepares features for
- Model training → Model evaluation: evaluated by
- Finance analyst → Flask application: submits request
- Flask application → Request preprocessing: calls
- Flask application → Loaded prediction model: loads and predicts with
- Request preprocessing → Loaded prediction model: supplies processed input
- Loaded prediction model → Sales prediction: returns prediction
- Flask application → Finance analyst: returns result
- Model training → LSTM forecasting: supports time-series methods