Merge pull request #71 from chenyangkang/beta

add ensemble_bootstrap; add backend argument passed to joblib; add base_model_method argument to predict method

Author: chenyangkang

docs/Examples/01.AdaSTEM_demo.ipynb

@@ -23,7 +23,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 1,
+   "execution_count": null,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -38,7 +38,6 @@
     "import matplotlib\n",
     "import warnings\n",
     "import pickle\n",
-    "import h3pandas\n",
     "\n",
     "pd.set_option('display.max_columns', None)\n",
     "# warnings.filterwarnings('ignore')\n"
@@ -731,7 +730,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 9,
+   "execution_count": null,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -741,6 +740,7 @@
     "        classifier=XGBClassifier(tree_method='hist',random_state=42, verbosity = 0, n_jobs=1),\n",
     "        regressor=XGBRegressor(tree_method='hist',random_state=42, verbosity = 0, n_jobs=1)\n",
     "    ),                                      # hurdel model for zero-inflated problem (e.g., count)\n",
+    "    task='hurdle',\n",
     "    save_gridding_plot = True,\n",
     "    ensemble_fold=10,                       # data are modeled 10 times, each time with jitter and rotation in Quadtree algo\n",
     "    min_ensemble_required=7,                # Only points covered by > 7 ensembles will be predicted\n",

docs/Examples/02.AdaSTEM_learning_curve_analysis.ipynb

@@ -136,7 +136,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 8,
+   "execution_count": null,
    "metadata": {},
    "outputs": [
     {
@@ -195,6 +195,7 @@
     "                classifier=XGBClassifier(tree_method='hist',random_state=42, verbosity = 0, n_jobs=1),\n",
     "                regressor=XGBRegressor(tree_method='hist',random_state=42, verbosity = 0, n_jobs=1)\n",
     "            ),\n",
+    "            task='hurdle',\n",
     "            save_gridding_plot = True,\n",
     "            ensemble_fold=10, \n",
     "            min_ensemble_required=7,\n",

docs/Examples/04.SphereAdaSTEM_demo.ipynb

@@ -721,7 +721,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 9,
+   "execution_count": null,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -731,6 +731,7 @@
     "        classifier=XGBClassifier(tree_method='hist',random_state=42, verbosity = 0, n_jobs=1),\n",
     "        regressor=XGBRegressor(tree_method='hist',random_state=42, verbosity = 0, n_jobs=1)\n",
     "    ),                                      # hurdel model for zero-inflated problem (e.g., count)\n",
+    "    task='hurdle',\n",
     "    save_gridding_plot = True,\n",
     "    ensemble_fold=10,                       # data are modeled 10 times, each time with jitter and rotation in Quadtree algo\n",
     "    min_ensemble_required=7,                # Only points covered by > 7 stixels will be predicted\n",

docs/Examples/05.Hurdle_in_ada_or_ada_in_hurdle.ipynb

@@ -651,7 +651,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 8,
+   "execution_count": null,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -660,6 +660,7 @@
     "        classifier=XGBClassifier(tree_method='hist',random_state=42, verbosity = 0, n_jobs=1),\n",
     "        regressor=XGBRegressor(tree_method='hist',random_state=42, verbosity = 0, n_jobs=1)\n",
     "    ),\n",
+    "    task='hurdle',\n",
     "    save_gridding_plot = True,\n",
     "    ensemble_fold=10, \n",
     "    min_ensemble_required=7,\n",

docs/Examples/06.Base_model_choices.ipynb

@@ -710,6 +710,7 @@
     "                classifier=base_model_dict[base_model_name]['classifier'],\n",
     "                regressor=base_model_dict[base_model_name]['regressor']\n",
     "            ),\n",
+    "            task='hurdle',\n",
     "            save_gridding_plot = True,\n",
     "            ensemble_fold=10, \n",
     "            min_ensemble_required=7,\n",

docs/Examples/07.Optimizing_stixel_size.ipynb

@@ -581,7 +581,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 10,
+   "execution_count": null,
    "metadata": {},
    "outputs": [
     {
@@ -1154,6 +1154,7 @@
     "        classifier=XGBClassifier(tree_method='hist',random_state=42, verbosity = 0, n_jobs=1),\n",
     "        regressor=XGBRegressor(tree_method='hist',random_state=42, verbosity = 0, n_jobs=1)\n",
     "    ),\n",
+    "    task='hurdle',\n",
     "    save_gridding_plot = True,\n",
     "    ensemble_fold=10, \n",
     "    min_ensemble_required=7,\n",

docs/Examples/08.Lazy_loading.ipynb

@@ -678,7 +678,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 9,
+   "execution_count": null,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -691,6 +691,7 @@
     "            classifier=XGBClassifier(tree_method='hist',random_state=42, verbosity = 0, n_jobs=1),\n",
     "            regressor=XGBRegressor(tree_method='hist',random_state=42, verbosity = 0, n_jobs=1)\n",
     "        ),                                      # hurdel model for zero-inflated problem (e.g., count)\n",
+    "        task='hurdle',\n",
     "        save_gridding_plot = True,\n",
     "        ensemble_fold=ensemble_fold,                       # data are modeled 10 times, each time with jitter and rotation in Quadtree algo\n",
     "        min_ensemble_required=ensemble_fold-2,                # Only points covered by > 7 ensembles will be predicted\n",
@@ -720,6 +721,7 @@
     "            classifier=XGBClassifier(tree_method='hist',random_state=42, verbosity = 0, n_jobs=1),\n",
     "            regressor=XGBRegressor(tree_method='hist',random_state=42, verbosity = 0, n_jobs=1)\n",
     "        ),                                      # hurdel model for zero-inflated problem (e.g., count)\n",
+    "        task='hurdle',\n",
     "        save_gridding_plot = True,\n",
     "        ensemble_fold=ensemble_fold,                       # data are modeled 10 times, each time with jitter and rotation in Quadtree algo\n",
     "        min_ensemble_required=ensemble_fold-2,                # Only points covered by > 7 ensembles will be predicted\n",