988 State Forecaster

Click a state marker on the map, or choose a state from the dropdown. The selected state is filtered from the CSV and modeled automatically.








For easiest use in XAMPP, put 988_state_monthly_panel_clean(5).csv in the same folder as this PHP file.

CA - California Forecast

Metric: answered
Selected model: OLS trend + monthly seasonality
Selection mode: auto
Selection reason: Automatic selection chose the model with the lowest pre-period RMSE.
Model reasoning: OLS fits a transparent linear trend plus repeating monthly seasonal effects.

RMSE: 897.99 | MAE: 781.73 | MAPE: 3.33%

Projection start: 2024-01-01

Model Comparison

Rank Model RMSE MAE MAPE Reasoning
1 OLS trend + monthly seasonality 897.99 781.73 3.33% OLS fits a transparent linear trend plus repeating monthly seasonal effects.
2 Prophet-like piecewise trend + monthly seasonality 1121.84 866.8 3.64% This pure PHP approximation mimics Prophet by allowing changepoints in the trend while preserving seasonality.
3 Theil-Sen robust trend + monthly seasonality 1211.6 756.88 3.38% Theil-Sen is more resistant to outliers and unusual spikes than OLS.
4 Holt-Winters additive seasonality 1691.63 1062.25 4.31% Holt-Winters captures changing level, trend, and recurring monthly seasonality.