Model Fit
Category: Mega-Polis → Analysis → Analysis Data Tools
Node ID:SvMegapolisModelFit
Tooltip: Model Fit
Dependencies:sklearn
Functionality
Fits a scikit-learn model using a train/test split.
When run is enabled, the node: 1. Splits X and y into training and test sets using sklearn.model_selection.train_test_split (with shuffle=True and the chosen train_size) 2. Fits the model on the training set (model.fit(X_train, y_train)) 3. Generates predictions for X_test (data.predict(X_test)) (predictions are computed but not output) 4. Outputs the fitted model instance as Model Out
When run is disabled, the node outputs an empty string.
Inputs
| Socket | Type | Description |
|---|---|---|
| Model | SvStringsSocket | A scikit-learn estimator instance (expected list-wrapped; the code uses model = self.model[0]). Must be linked. |
| X | SvStringsSocket | Feature matrix for training/testing. Must be linked. |
| y | SvStringsSocket | Target values. Must be linked. |
Parameters
| Name | Type | Default | Description |
|---|---|---|---|
| run | Bool | False | When enabled, performs the train/test split and fits the model. |
| train_size | Float | 0.30 | Proportion of samples used for training in train_test_split(...) (min 0.01, max 0.99). |
Outputs
| Socket | Type | Description |
|---|---|---|
| Model Out | SvStringsSocket | The fitted model instance wrapped as a list: [data] when run is enabled; otherwise an empty string. |
Example
Fit a model (basic workflow)
- Use Linear Model Selection (or any other model factory node) to create a model.
- Prepare:
- X (feature matrix)
- y (target vector)
- Connect:
- Model → Model
- X → X
- y → y
- Enable run and set train_size (e.g.,
0.70for 70% training). - Use Model Out as input to your Model Predict / Model Evaluate nodes.