BasePipeline class
==================

.. class:: BasePipeline(self, model: sam_ml.models.main_classifier.Classifier | sam_ml.models.main_regressor.Regressor, vectorizer: str | sam_ml.data.preprocessing.embeddings.Embeddings_builder | None, scaler: str | sam_ml.data.preprocessing.scaler.Scaler | None, selector: str | tuple[str, int] | sam_ml.data.preprocessing.feature_selection.Selector | None, sampler: str | sam_ml.data.preprocessing.sampling.Sampler | sam_ml.data.preprocessing.sampling_pipeline.SamplerPipeline | None, model_name: str)

BasePipeline class  - parent class :class:`Model`

.. list-table::
   :widths: 25 75
   :header-rows: 0

   * - Parameters
     - 
        model : Classifier or Regressor class object
            Model used in pipeline (:class:`Classifier` or :class:`Regressor`)
        vectorizer : str, Embeddings_builder, or None
            object or algorithm of :class:`Embeddings_builder` class which will be used for automatic string column vectorizing (None for no vectorizing)
        scaler : str, Scaler, or None
            object or algorithm of :class:`Scaler` class for scaling the data (None for no scaling)
        selector : str, Selector, or None
            object, tuple of algorithm and feature number, or algorithm of :class:`Selector` class for feature selection (None for no selecting)
        sampler : str, Sampler, SamplerPipeline, or None
            object or algorithm of :class:`Sampler` / :class:`SamplerPipeline` class for sampling the train data (None for no sampling)
        model_name : str
            name of the model
        

   * - Attributes
     - 
         
        cv_scores : dict[str, float]
            dictionary with cross validation results
        
       
         
        data_classes_trained : bool
            If ``True``, the preprocessing step classes are fitted. Important for methods that use warm_start
        
       
         
        feature_names : list[str]
            names of all the features that the model saw during training. Is empty if model was not fitted yet.
        
       
         
        grid : ConfigurationSpace
            hyperparameter tuning grid of the model
        
       
         
        model : model object
            model with 'fit', 'predict', 'set_params', and 'get_params' method (see sklearn API)
        
       
         
        model_name : str
            name of the model. Used in loading bars and dictionaries as identifier of the model
        
       
         
        model_type : str
            kind of estimator (e.g. 'RFC' for RandomForestClassifier)
        
       
         
        rCVsearch_results : pd.DataFrame or None
            results from randomCV hyperparameter tuning. Is ``None`` if randomCVsearch was not used yet.
        
       
         
        steps : list[tuple[str, any]]
            list with preprocessing + model pipeline steps as tuples
        
       
         
        string_columns : list[str]
            list with detected string columns that are used in auto-vectorizing
        
       
         
        train_score : float
            train score value
        
       
         
        train_time : str
            train time in format: "0:00:00" (hours:minutes:seconds)
        
       





.. raw:: html

   <h2>Methods</h2>

.. list-table::
   :widths: 25 75
   :header-rows: 1

   * - Method
     - Description
   
   * - :meth:`~sam_ml.models.main_pipeline.BasePipeline._auto_vectorizing`
     - Function to detect string columns and creating a vectorizer for each, and vectorize them
   
   * - :meth:`~sam_ml.models.main_pipeline.BasePipeline._changed_parameters`
     - Function to get parameters that differ from the default ones
   
   * - :meth:`~sam_ml.models.main_pipeline.BasePipeline._data_prepare`
     - Function to run data class objects on data to prepare them for the model
   
   * - :meth:`~sam_ml.models.main_pipeline.BasePipeline._get_all_scores`
     - Function to create multiple scores for given y_true-y_pred pairs
   
   * - :meth:`~sam_ml.models.main_pipeline.BasePipeline._get_score`
     - Calculate a score for given y true and y prediction values
   
   * - :meth:`~sam_ml.models.main_pipeline.BasePipeline._inherit_from_model`
     - Function to inherit methods and attributes from model
   
   * - :meth:`~sam_ml.models.main_pipeline.BasePipeline._make_cv_scores`
     - Function to create from the crossvalidation results a dictionary
   
   * - :meth:`~sam_ml.models.main_pipeline.BasePipeline._make_scorer`
     - Function to create a dictionary with scorer for the crossvalidation
   
   * - :meth:`~sam_ml.models.main_pipeline.BasePipeline._print_scores`
     - Function to print out the values of a dictionary
   
   * - :meth:`~sam_ml.models.main_pipeline.BasePipeline._validate_component`
     - Function to create the data preprocessing steps
   
   * - :meth:`~sam_ml.models.main_pipeline.BasePipeline.cross_validation`
     - Random split crossvalidation
   
   * - :meth:`~sam_ml.models.main_pipeline.BasePipeline.cross_validation_small_data`
     - One-vs-all cross validation for small datasets
   
   * - :meth:`~sam_ml.models.main_pipeline.BasePipeline.evaluate`
     - Function to create multiple scores with predict function of model
   
   * - :meth:`~sam_ml.models.main_pipeline.BasePipeline.evaluate_score`
     - Function to create a score with self.__get_score of model
   
   * - :meth:`~sam_ml.models.main_pipeline.BasePipeline.feature_importance`
     - Function to generate a matplotlib plot of the top45 feature importance from the model.
   
   * - :meth:`~sam_ml.models.main_pipeline.BasePipeline.fit`
     - Function to fit the model
   
   * - :meth:`~sam_ml.models.main_pipeline.BasePipeline.fit_warm_start`
     - Function to warm_start fit the model
   
   * - :meth:`~sam_ml.models.main_pipeline.BasePipeline.get_deepcopy`
     - Function to create a deepcopy of object
   
   * - :meth:`~sam_ml.models.main_pipeline.BasePipeline.get_params`
     - Function to get the parameter from the model object
   
   * - :meth:`~sam_ml.models.main_pipeline.BasePipeline.get_random_config`
     - Function to generate one grid configuration
   
   * - :meth:`~sam_ml.models.main_pipeline.BasePipeline.get_random_configs`
     - Function to generate grid configurations
   
   * - :meth:`~sam_ml.models.main_pipeline.BasePipeline.load_model`
     - Function to load a pickled model class object
   
   * - :meth:`~sam_ml.models.main_pipeline.BasePipeline.predict`
     - Function to predict with predict-method from model object
   
   * - :meth:`~sam_ml.models.main_pipeline.BasePipeline.predict_proba`
     - Function to predict with predict_proba-method from model object
   
   * - :meth:`~sam_ml.models.main_pipeline.BasePipeline.randomCVsearch`
     - Hyperparametertuning with randomCVsearch
   
   * - :meth:`~sam_ml.models.main_pipeline.BasePipeline.replace_grid`
     - Function to replace self.grid
   
   * - :meth:`~sam_ml.models.main_pipeline.BasePipeline.save_model`
     - Function to pickle and save the class object
   
   * - :meth:`~sam_ml.models.main_pipeline.BasePipeline.set_params`
     - Function to set the parameter of the model object
   
   * - :meth:`~sam_ml.models.main_pipeline.BasePipeline.smac_search`
     - Hyperparametertuning with SMAC library HyperparameterOptimizationFacade [can only be used in the sam_ml version with swig]
   
   * - :meth:`~sam_ml.models.main_pipeline.BasePipeline.train`
     - Function to train the model
   
   * - :meth:`~sam_ml.models.main_pipeline.BasePipeline.train_warm_start`
     - Function to warm_start train the model
   




.. automethod:: sam_ml.models.main_pipeline.BasePipeline._auto_vectorizing

.. automethod:: sam_ml.models.main_pipeline.BasePipeline._changed_parameters

.. automethod:: sam_ml.models.main_pipeline.BasePipeline._data_prepare

.. automethod:: sam_ml.models.main_pipeline.BasePipeline._get_all_scores

.. automethod:: sam_ml.models.main_pipeline.BasePipeline._get_score

.. automethod:: sam_ml.models.main_pipeline.BasePipeline._inherit_from_model

.. automethod:: sam_ml.models.main_pipeline.BasePipeline._make_cv_scores

.. automethod:: sam_ml.models.main_pipeline.BasePipeline._make_scorer

.. automethod:: sam_ml.models.main_pipeline.BasePipeline._print_scores

.. automethod:: sam_ml.models.main_pipeline.BasePipeline._validate_component

.. automethod:: sam_ml.models.main_pipeline.BasePipeline.cross_validation

.. automethod:: sam_ml.models.main_pipeline.BasePipeline.cross_validation_small_data

.. automethod:: sam_ml.models.main_pipeline.BasePipeline.evaluate

.. automethod:: sam_ml.models.main_pipeline.BasePipeline.evaluate_score

.. automethod:: sam_ml.models.main_pipeline.BasePipeline.feature_importance

.. automethod:: sam_ml.models.main_pipeline.BasePipeline.fit

.. automethod:: sam_ml.models.main_pipeline.BasePipeline.fit_warm_start

.. automethod:: sam_ml.models.main_pipeline.BasePipeline.get_deepcopy

.. automethod:: sam_ml.models.main_pipeline.BasePipeline.get_params

.. automethod:: sam_ml.models.main_pipeline.BasePipeline.get_random_config

.. automethod:: sam_ml.models.main_pipeline.BasePipeline.get_random_configs

.. automethod:: sam_ml.models.main_pipeline.BasePipeline.load_model

.. automethod:: sam_ml.models.main_pipeline.BasePipeline.predict

.. automethod:: sam_ml.models.main_pipeline.BasePipeline.predict_proba

.. automethod:: sam_ml.models.main_pipeline.BasePipeline.randomCVsearch

.. automethod:: sam_ml.models.main_pipeline.BasePipeline.replace_grid

.. automethod:: sam_ml.models.main_pipeline.BasePipeline.save_model

.. automethod:: sam_ml.models.main_pipeline.BasePipeline.set_params

.. automethod:: sam_ml.models.main_pipeline.BasePipeline.smac_search

.. automethod:: sam_ml.models.main_pipeline.BasePipeline.train

.. automethod:: sam_ml.models.main_pipeline.BasePipeline.train_warm_start