added wSAA based on RF_LGBM but boosting_type has still be specified …

…manually

_proc/01_levelSetKDEx_univariate.ipynb

@@ -816,7 +816,6 @@
    ]
   },
   {
-   "attachments": {},
    "cell_type": "markdown",
    "id": "fcdc893f",
    "metadata": {},
@@ -1373,9 +1372,9 @@
  ],
  "metadata": {
   "kernelspec": {
-   "display_name": "python3",
+   "display_name": "dddex",
    "language": "python",
-   "name": "python3"
+   "name": "dddex"
   }
  },
  "nbformat": 4,

_proc/03_wSAA.ipynb

@@ -27,7 +27,16 @@
    "metadata": {
     "language": "python"
    },
-   "outputs": [],
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "The autoreload extension is already loaded. To reload it, use:\n",
+      "  %reload_ext autoreload\n"
+     ]
+    }
+   ],
    "source": []
   },
   {
@@ -72,7 +81,7 @@
       "text/markdown": [
        "---\n",
        "\n",
-       "[source](https://github.com/kaiguender/dddex/blob/main/dddex/wSAA.py#L22){target=\"_blank\" style=\"float:right; font-size:smaller\"}\n",
+       "[source](https://github.com/kaiguender/dddex/blob/main/dddex/wSAA.py#L23){target=\"_blank\" style=\"float:right; font-size:smaller\"}\n",
        "\n",
        "### RandomForestWSAA\n",
        "\n",
@@ -119,7 +128,7 @@
       "text/plain": [
        "---\n",
        "\n",
-       "[source](https://github.com/kaiguender/dddex/blob/main/dddex/wSAA.py#L22){target=\"_blank\" style=\"float:right; font-size:smaller\"}\n",
+       "[source](https://github.com/kaiguender/dddex/blob/main/dddex/wSAA.py#L23){target=\"_blank\" style=\"float:right; font-size:smaller\"}\n",
        "\n",
        "### RandomForestWSAA\n",
        "\n",
@@ -211,6 +220,83 @@
     "# show_doc(RandomForestWSAA.getWeights)"
    ]
   },
+  {
+   "cell_type": "markdown",
+   "id": "2811aafa",
+   "metadata": {},
+   "source": [
+    "## wSAA - Random Forest LightGBM"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 2,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/markdown": [
+       "---\n",
+       "\n",
+       "[source](https://github.com/kaiguender/dddex/blob/main/dddex/wSAA.py#L113){target=\"_blank\" style=\"float:right; font-size:smaller\"}\n",
+       "\n",
+       "### RandomForestWSAA_LGBM\n",
+       "\n",
+       ">      RandomForestWSAA_LGBM (boosting_type:str='gbdt', num_leaves:int=31,\n",
+       ">                             max_depth:int=-1, learning_rate:float=0.1,\n",
+       ">                             n_estimators:int=100,\n",
+       ">                             subsample_for_bin:int=200000,\n",
+       ">                             objective:Union[str,Callable,NoneType]=None,\n",
+       ">                             class_weight:Union[Dict,str,NoneType]=None,\n",
+       ">                             min_split_gain:float=0.0,\n",
+       ">                             min_child_weight:float=0.001,\n",
+       ">                             min_child_samples:int=20, subsample:float=1.0,\n",
+       ">                             subsample_freq:int=0, colsample_bytree:float=1.0,\n",
+       ">                             reg_alpha:float=0.0, reg_lambda:float=0.0, random_\n",
+       ">                             state:Union[int,numpy.random.mtrand.RandomState,No\n",
+       ">                             neType]=None, n_jobs:int=-1,\n",
+       ">                             silent:Union[bool,str]='warn',\n",
+       ">                             importance_type:str='split', **kwargs)\n",
+       "\n",
+       "LightGBM regressor."
+      ],
+      "text/plain": [
+       "---\n",
+       "\n",
+       "[source](https://github.com/kaiguender/dddex/blob/main/dddex/wSAA.py#L113){target=\"_blank\" style=\"float:right; font-size:smaller\"}\n",
+       "\n",
+       "### RandomForestWSAA_LGBM\n",
+       "\n",
+       ">      RandomForestWSAA_LGBM (boosting_type:str='gbdt', num_leaves:int=31,\n",
+       ">                             max_depth:int=-1, learning_rate:float=0.1,\n",
+       ">                             n_estimators:int=100,\n",
+       ">                             subsample_for_bin:int=200000,\n",
+       ">                             objective:Union[str,Callable,NoneType]=None,\n",
+       ">                             class_weight:Union[Dict,str,NoneType]=None,\n",
+       ">                             min_split_gain:float=0.0,\n",
+       ">                             min_child_weight:float=0.001,\n",
+       ">                             min_child_samples:int=20, subsample:float=1.0,\n",
+       ">                             subsample_freq:int=0, colsample_bytree:float=1.0,\n",
+       ">                             reg_alpha:float=0.0, reg_lambda:float=0.0, random_\n",
+       ">                             state:Union[int,numpy.random.mtrand.RandomState,No\n",
+       ">                             neType]=None, n_jobs:int=-1,\n",
+       ">                             silent:Union[bool,str]='warn',\n",
+       ">                             importance_type:str='split', **kwargs)\n",
+       "\n",
+       "LightGBM regressor."
+      ]
+     },
+     "execution_count": 2,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "#| echo: false\n",
+    "#| output: asis\n",
+    "show_doc(RandomForestWSAA_LGBM)"
+   ]
+  },
   {
    "cell_type": "markdown",
    "id": "f471bba6-50ec-49a9-980d-1c10f4361938",
@@ -221,15 +307,15 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 2,
+   "execution_count": 3,
    "metadata": {},
    "outputs": [
     {
      "data": {
       "text/markdown": [
        "---\n",
        "\n",
-       "[source](https://github.com/kaiguender/dddex/blob/main/dddex/wSAA.py#L112){target=\"_blank\" style=\"float:right; font-size:smaller\"}\n",
+       "[source](https://github.com/kaiguender/dddex/blob/main/dddex/wSAA.py#L203){target=\"_blank\" style=\"float:right; font-size:smaller\"}\n",
        "\n",
        "### SampleAverageApproximation\n",
        "\n",
@@ -241,7 +327,7 @@
       "text/plain": [
        "---\n",
        "\n",
-       "[source](https://github.com/kaiguender/dddex/blob/main/dddex/wSAA.py#L112){target=\"_blank\" style=\"float:right; font-size:smaller\"}\n",
+       "[source](https://github.com/kaiguender/dddex/blob/main/dddex/wSAA.py#L203){target=\"_blank\" style=\"float:right; font-size:smaller\"}\n",
        "\n",
        "### SampleAverageApproximation\n",
        "\n",
@@ -251,7 +337,7 @@
        "by assigning equal probability to each historical observation of said target variable."
       ]
      },
-     "execution_count": 2,
+     "execution_count": 3,
      "metadata": {},
      "output_type": "execute_result"
     }
@@ -325,13 +411,53 @@
    "source": [
     "# Test Code"
    ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "c5f055de-89ce-4943-8242-8f434ee8a3f1",
+   "metadata": {
+    "language": "python"
+   },
+   "outputs": [],
+   "source": [
+    "# #| hide\n",
+    "\n",
+    "# from lightgbm import LGBMRegressor\n",
+    "# import lightgbm as lgb\n",
+    "# from dddex.loadData import *\n",
+    "# import ipdb\n",
+    "# import inspect\n",
+    "# from sklearn.base import RegressorMixin\n",
+    "\n",
+    "# data, XTrain, yTrain, XTest, yTest = loadDataBakery()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "bbcd2880-1b9e-4d07-ab84-4ba8a2bef226",
+   "metadata": {
+    "language": "python"
+   },
+   "outputs": [],
+   "source": [
+    "# #| hide\n",
+    "\n",
+    "# RF = RandomForestWSAA_LGBM(max_depth = 2,\n",
+    "#                            n_estimators = 10,\n",
+    "#                            n_jobs = 1,\n",
+    "#                            boosting_type = 'rf',\n",
+    "#                            subsample_freq = 0,\n",
+    "#                            subsample = 1)"
+   ]
   }
  ],
  "metadata": {
   "kernelspec": {
-   "display_name": "dddex",
+   "display_name": "python3",
    "language": "python",
-   "name": "dddex"
+   "name": "python3"
   }
  },
  "nbformat": 4,

dddex/_modidx.py

@@ -140,6 +140,12 @@
                             'dddex.wSAA.RandomForestWSAA.getWeights': ('wsaa.html#randomforestwsaa.getweights', 'dddex/wSAA.py'),
                             'dddex.wSAA.RandomForestWSAA.pointPredict': ('wsaa.html#randomforestwsaa.pointpredict', 'dddex/wSAA.py'),
                             'dddex.wSAA.RandomForestWSAA.predict': ('wsaa.html#randomforestwsaa.predict', 'dddex/wSAA.py'),
+                            'dddex.wSAA.RandomForestWSAA_LGBM': ('wsaa.html#randomforestwsaa_lgbm', 'dddex/wSAA.py'),
+                            'dddex.wSAA.RandomForestWSAA_LGBM.fit': ('wsaa.html#randomforestwsaa_lgbm.fit', 'dddex/wSAA.py'),
+                            'dddex.wSAA.RandomForestWSAA_LGBM.getWeights': ('wsaa.html#randomforestwsaa_lgbm.getweights', 'dddex/wSAA.py'),
+                            'dddex.wSAA.RandomForestWSAA_LGBM.pointPredict': ( 'wsaa.html#randomforestwsaa_lgbm.pointpredict',
+                                                                               'dddex/wSAA.py'),
+                            'dddex.wSAA.RandomForestWSAA_LGBM.predict': ('wsaa.html#randomforestwsaa_lgbm.predict', 'dddex/wSAA.py'),
                             'dddex.wSAA.SampleAverageApproximation': ('wsaa.html#sampleaverageapproximation', 'dddex/wSAA.py'),
                             'dddex.wSAA.SampleAverageApproximation.__init__': ( 'wsaa.html#sampleaverageapproximation.__init__',
                                                                                 'dddex/wSAA.py'),

dddex/wSAA.py

@@ -11,12 +11,13 @@
 import copy
 
 from sklearn.ensemble import RandomForestRegressor
+from lightgbm import LGBMRegressor
 from sklearn.base import MetaEstimatorMixin
 from .baseClasses import BaseWeightsBasedEstimator
 from .utils import restructureWeightsDataList
 
 # %% auto 0
-__all__ = ['RandomForestWSAA', 'SampleAverageApproximation']
+__all__ = ['RandomForestWSAA', 'RandomForestWSAA_LGBM', 'SampleAverageApproximation']
 
 # %% ../nbs/03_wSAA.ipynb 7
 class RandomForestWSAA(RandomForestRegressor, BaseWeightsBasedEstimator):
@@ -109,6 +110,96 @@ def pointPredict(self,
 
 
 # %% ../nbs/03_wSAA.ipynb 12
+class RandomForestWSAA_LGBM(LGBMRegressor, BaseWeightsBasedEstimator):
+
+    def fit(self, 
+            X: np.ndarray, # Feature matrix
+            y: np.ndarray, # Target values
+            **kwargs):
+
+        super().fit(X = X, 
+                    y = y, 
+                    **kwargs)
+
+        self.yTrain = y
+
+        self.leafIndicesTrain = self.pointPredict(X, pred_leaf = True)
+
+    #---
+
+    def getWeights(self, 
+                   X: np.ndarray, # Feature matrix for which conditional density estimates are computed.
+                   # Specifies structure of the returned density estimates. One of: 
+                   # 'all', 'onlyPositiveWeights', 'summarized', 'cumDistribution', 'cumDistributionSummarized'
+                   outputType: str='onlyPositiveWeights', 
+                   # Optional. List with length X.shape[0]. Values are multiplied to the estimated 
+                   # density of each sample for scaling purposes.
+                   scalingList: list=None, 
+                   ) -> list: # List whose elements are the conditional density estimates for the samples specified by `X`.
+
+        __doc__ = BaseWeightsBasedEstimator.getWeights.__doc__
+
+        #---
+
+        leafIndicesDf = self.pointPredict(X, pred_leaf = True)
+
+        weightsDataList = list()
+
+        for leafIndices in leafIndicesDf:
+            leafComparisonMatrix = (self.leafIndicesTrain == leafIndices) * 1
+            nObsInSameLeaf = np.sum(leafComparisonMatrix, axis = 0)
+
+            # It can happen that RF decides that the best strategy is to fit no tree at
+            # all and simply average all results (happens when min_child_sample is too high, for example).
+            # In this case 'leafComparisonMatrix' mustn't be averaged because there has been only a single tree.
+            if len(leafComparisonMatrix.shape) == 1:
+                weights = leafComparisonMatrix / nObsInSameLeaf
+            else:
+                weights = np.mean(leafComparisonMatrix / nObsInSameLeaf, axis = 1)
+
+            weightsPosIndex = np.where(weights > 0)[0]
+
+            weightsDataList.append((weights[weightsPosIndex], weightsPosIndex))
+
+        #---
+
+        weightsDataList = restructureWeightsDataList(weightsDataList = weightsDataList, 
+                                                     outputType = outputType, 
+                                                     y = self.yTrain, 
+                                                     scalingList = scalingList,
+                                                     equalWeights = False)
+
+        return weightsDataList
+
+    #---
+
+    def predict(self : BaseWeightsBasedEstimator, 
+                X: np.ndarray, # Feature matrix for which conditional quantiles are computed.
+                probs: list, # Probabilities for which quantiles are computed.
+                outputAsDf: bool=True, # Determines output. Either a dataframe with probs as columns or a dict with probs as keys.
+                # Optional. List with length X.shape[0]. Values are multiplied to the predictions
+                # of each sample to rescale values.
+                scalingList: list=None, 
+                ): 
+
+        __doc__ = BaseWeightsBasedEstimator.predict.__doc__
+
+        return super(MetaEstimatorMixin, self).predict(X = X,
+                                                       probs = probs, 
+                                                       scalingList = scalingList)
+
+    #---
+
+    def pointPredict(self,
+                     X: np.ndarray, # Feature Matrix
+                     **kwargs):
+        """Original `predict` method to generate point forecasts"""
+
+        return super().predict(X = X,
+                               **kwargs)
+
+
+# %% ../nbs/03_wSAA.ipynb 14
 class SampleAverageApproximation(BaseWeightsBasedEstimator):
     """SAA is a featureless approach that assumes the density of the target variable is given
     by assigning equal probability to each historical observation of said target variable."""