Predict - Prediction Engine

<div class="section-label">03 Predict</div> <div class="section-title">Partial-sample regression at thresholds you choose.</div> Predict forms a relevance-weighted forecast at the censoring threshold(s) you choose. That is **partial-sample regression**: the prediction uses the informative, relevant subset of the sample, rather than treating every past observation equally. It is the evaluation inside each [[Grid Prediction|Grid]] cell and inside [[MaxFit]]. Use Predict when the attribute set and thresholds are already fixed. For primary analysis, prefer [[Grid Prediction]]. Takes the same inputs as the other prediction calls: outcomes y, attributes X, and circumstances theta. ## Signature ```text predict( y, # outcomes, length N X, # attributes, N × K theta, # circumstances, length K options? # PredictOptions; omit → defaults ) → results ``` **Call:** `predict` **Options:** `PredictOptions` **Returns:** [[Results/Overview|Results]] ## Arguments | Name | Shape | Description | | --- | --- | --- | | y | N | Observed outcomes | | X | N × K | Training attributes aligned with y | | theta | K | Circumstances for the prediction task | | options | (object) | PredictOptions; optional, omit for defaults | N = number of observations (rows), K = number of independent variables (columns). ## Options A Predict call is ready with defaults: threshold [0.5] and censor_type relevance. Pass a [[Settings/Options#PredictOptions|PredictOptions]] object when you want to change those, or any other Predict setting. Predict censors on **relevance** or **similarity**, not both at once. To compare both in one run, use [[MaxFit]] or [[Grid Prediction]]. The full parameter list and allowed value names: [[Settings/Options#PredictOptions|PredictOptions]] · [[Settings/Allowed Values|Allowed values]]. ## Results A successful Predict call gives you a forecast at the threshold(s) you chose, with fit alongside it, plus observation weights and relevance scores for that calibration. | You get | Look here | | --- | --- | | Forecast and fit | yhat, fit, adjusted_fit, … — [[Results/Core\|Core]] | | Weights and scores | prediction_weights, insights — [[Results/Weights & Insights\|Weights & Insights]] | | Solo distribution | solo_distribution — [[Results/Solo Distribution\|Solo Distribution]] | The full layout of a result object: [[Results/Overview|Results]]. ## Related - [[Functions/Overview|Functions]] - [[Grid Prediction]] · [[MaxFit]] - [[Settings/Options#PredictOptions|PredictOptions]] · [[Results/Overview|Results]]