Pipeline(steps=[('missing', MissingValueThreshold()),\n",
+ "Pipeline(steps=[('missing', MissingValueThreshold()),\n",
" ('unique', UniqueValuesThreshold()),\n",
" ('cardinality', CardinalityThreshold(threshold=10)),\n",
" ('collinearity', CollinearityThreshold(threshold=0.75)),\n",
@@ -2459,7 +2447,7 @@
" 'objective': 'rmse',\n",
" 'verbosity': -1,\n",
" 'zero_as_missing': False},\n",
- " verbose=2))]) In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org. Pipeline Pipeline(steps=[('missing', MissingValueThreshold()),\n",
+ " verbose=2))]) In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org. Pipeline Pipeline(steps=[('missing', MissingValueThreshold()),\n",
" ('unique', UniqueValuesThreshold()),\n",
" ('cardinality', CardinalityThreshold(threshold=10)),\n",
" ('collinearity', CollinearityThreshold(threshold=0.75)),\n",
@@ -2470,7 +2458,7 @@
" 'objective': 'rmse',\n",
" 'verbosity': -1,\n",
" 'zero_as_missing': False},\n",
- " verbose=2))]) "
@@ -2525,9 +2513,9 @@
"
\n",
" \n",
" \n",
+ " var1 \n",
" var3 \n",
" var5 \n",
- " var6 \n",
" var8 \n",
" var9 \n",
" \n",
@@ -2535,41 +2523,41 @@
" \n",
" \n",
" 0 \n",
+ " 0.821664 \n",
" 0.985738 \n",
" -0.494190 \n",
- " 2 \n",
" -0.361717 \n",
" 1 \n",
" \n",
" \n",
" 1 \n",
+ " 0.334013 \n",
" 0.819273 \n",
" 0.283845 \n",
- " 1 \n",
" 0.178670 \n",
" 2 \n",
" \n",
" \n",
" 2 \n",
+ " 0.187234 \n",
" 1.087443 \n",
" 0.962503 \n",
- " 2 \n",
" -3.375579 \n",
" 2 \n",
" \n",
" \n",
" 3 \n",
+ " 0.994528 \n",
" 1.592398 \n",
" 1.165595 \n",
- " 0 \n",
" -0.449650 \n",
" 2 \n",
" \n",
" \n",
" 4 \n",
+ " 0.184859 \n",
" 0.865702 \n",
" -0.058833 \n",
- " 1 \n",
" 0.763903 \n",
" 1 \n",
" \n",
@@ -2578,12 +2566,12 @@
""
],
"text/plain": [
- " var3 var5 var6 var8 var9\n",
- "0 0.985738 -0.494190 2 -0.361717 1\n",
- "1 0.819273 0.283845 1 0.178670 2\n",
- "2 1.087443 0.962503 2 -3.375579 2\n",
- "3 1.592398 1.165595 0 -0.449650 2\n",
- "4 0.865702 -0.058833 1 0.763903 1"
+ " var1 var3 var5 var8 var9\n",
+ "0 0.821664 0.985738 -0.494190 -0.361717 1\n",
+ "1 0.334013 0.819273 0.283845 0.178670 2\n",
+ "2 0.187234 1.087443 0.962503 -3.375579 2\n",
+ "3 0.994528 1.592398 1.165595 -0.449650 2\n",
+ "4 0.184859 0.865702 -0.058833 0.763903 1"
]
},
"execution_count": 34,
@@ -2625,176 +2613,176 @@
"data": {
"text/html": [
"\n",
- "\n",
+ "\n",
" \n",
" \n",
" \n",
- " predictor \n",
- " missing \n",
- " unique \n",
- " cardinality \n",
- " collinearity \n",
- " encoder \n",
- " lowimp \n",
+ " predictor \n",
+ " missing \n",
+ " unique \n",
+ " cardinality \n",
+ " collinearity \n",
+ " encoder \n",
+ " lowimp \n",
" \n",
" \n",
" \n",
" \n",
- " 0 \n",
- " var0 \n",
- " 1 \n",
- " 1 \n",
- " 1 \n",
- " 0 \n",
- " nan \n",
- " nan \n",
- " \n",
- " \n",
- " 1 \n",
- " var1 \n",
- " 1 \n",
- " 1 \n",
- " 1 \n",
- " 0 \n",
- " nan \n",
- " nan \n",
- " \n",
- " \n",
- " 2 \n",
- " var2 \n",
- " 1 \n",
- " 1 \n",
- " 1 \n",
- " 0 \n",
- " nan \n",
- " nan \n",
- " \n",
- " \n",
- " 3 \n",
- " var3 \n",
- " 1 \n",
- " 1 \n",
- " 1 \n",
- " 1 \n",
- " nan \n",
- " 1 \n",
- " \n",
- " \n",
- " 4 \n",
- " var4 \n",
- " 1 \n",
- " 1 \n",
- " 1 \n",
- " 0 \n",
- " nan \n",
- " nan \n",
- " \n",
- " \n",
- " 5 \n",
- " var5 \n",
- " 1 \n",
- " 1 \n",
- " 1 \n",
- " 1 \n",
- " nan \n",
- " 1 \n",
- " \n",
- " \n",
- " 6 \n",
- " var6 \n",
- " 1 \n",
- " 1 \n",
- " 1 \n",
- " 1 \n",
- " nan \n",
- " 1 \n",
- " \n",
- " \n",
- " 7 \n",
- " var7 \n",
- " 1 \n",
- " 1 \n",
- " 1 \n",
- " 1 \n",
- " nan \n",
- " 0 \n",
- " \n",
- " \n",
- " 8 \n",
- " var8 \n",
- " 1 \n",
- " 1 \n",
- " 1 \n",
- " 1 \n",
- " nan \n",
- " 1 \n",
- " \n",
- " \n",
- " 9 \n",
- " var9 \n",
- " 1 \n",
- " 1 \n",
- " 1 \n",
- " 1 \n",
- " nan \n",
- " 1 \n",
- " \n",
- " \n",
- " 10 \n",
- " var10 \n",
- " 1 \n",
- " 0 \n",
- " nan \n",
- " nan \n",
- " nan \n",
- " nan \n",
- " \n",
- " \n",
- " 11 \n",
- " var11 \n",
- " 1 \n",
- " 1 \n",
- " 0 \n",
- " nan \n",
- " nan \n",
- " nan \n",
- " \n",
- " \n",
- " 12 \n",
- " var12 \n",
- " 0 \n",
- " nan \n",
- " nan \n",
- " nan \n",
- " nan \n",
- " nan \n",
- " \n",
- " \n",
- " 13 \n",
- " nice_guys \n",
- " 1 \n",
- " 1 \n",
- " 0 \n",
- " nan \n",
- " nan \n",
- " nan \n",
+ " 0 \n",
+ " var0 \n",
+ " 1 \n",
+ " 1 \n",
+ " 1 \n",
+ " 0 \n",
+ " nan \n",
+ " nan \n",
+ " \n",
+ " \n",
+ " 1 \n",
+ " var1 \n",
+ " 1 \n",
+ " 1 \n",
+ " 1 \n",
+ " 1 \n",
+ " nan \n",
+ " 1 \n",
+ " \n",
+ " \n",
+ " 2 \n",
+ " var2 \n",
+ " 1 \n",
+ " 1 \n",
+ " 1 \n",
+ " 0 \n",
+ " nan \n",
+ " nan \n",
+ " \n",
+ " \n",
+ " 3 \n",
+ " var3 \n",
+ " 1 \n",
+ " 1 \n",
+ " 1 \n",
+ " 1 \n",
+ " nan \n",
+ " 1 \n",
+ " \n",
+ " \n",
+ " 4 \n",
+ " var4 \n",
+ " 1 \n",
+ " 1 \n",
+ " 1 \n",
+ " 0 \n",
+ " nan \n",
+ " nan \n",
+ " \n",
+ " \n",
+ " 5 \n",
+ " var5 \n",
+ " 1 \n",
+ " 1 \n",
+ " 1 \n",
+ " 1 \n",
+ " nan \n",
+ " 1 \n",
+ " \n",
+ " \n",
+ " 6 \n",
+ " var6 \n",
+ " 1 \n",
+ " 1 \n",
+ " 1 \n",
+ " 1 \n",
+ " nan \n",
+ " 0 \n",
+ " \n",
+ " \n",
+ " 7 \n",
+ " var7 \n",
+ " 1 \n",
+ " 1 \n",
+ " 1 \n",
+ " 1 \n",
+ " nan \n",
+ " 0 \n",
+ " \n",
+ " \n",
+ " 8 \n",
+ " var8 \n",
+ " 1 \n",
+ " 1 \n",
+ " 1 \n",
+ " 1 \n",
+ " nan \n",
+ " 1 \n",
+ " \n",
+ " \n",
+ " 9 \n",
+ " var9 \n",
+ " 1 \n",
+ " 1 \n",
+ " 1 \n",
+ " 1 \n",
+ " nan \n",
+ " 1 \n",
+ " \n",
+ " \n",
+ " 10 \n",
+ " var10 \n",
+ " 1 \n",
+ " 0 \n",
+ " nan \n",
+ " nan \n",
+ " nan \n",
+ " nan \n",
+ " \n",
+ " \n",
+ " 11 \n",
+ " var11 \n",
+ " 1 \n",
+ " 1 \n",
+ " 0 \n",
+ " nan \n",
+ " nan \n",
+ " nan \n",
+ " \n",
+ " \n",
+ " 12 \n",
+ " var12 \n",
+ " 0 \n",
+ " nan \n",
+ " nan \n",
+ " nan \n",
+ " nan \n",
+ " nan \n",
+ " \n",
+ " \n",
+ " 13 \n",
+ " nice_guys \n",
+ " 1 \n",
+ " 1 \n",
+ " 0 \n",
+ " nan \n",
+ " nan \n",
+ " nan \n",
" \n",
" \n",
"
\n"
],
"text/plain": [
- ""
+ ""
]
},
"execution_count": 36,
@@ -2823,7 +2811,7 @@
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
- "version": "3.10.12"
+ "version": "3.10.0"
},
"vscode": {
"interpreter": {
diff --git a/docs/notebooks/mrmr_feature_selection.ipynb b/docs/notebooks/mrmr_feature_selection.ipynb
index 5415360..3659598 100644
--- a/docs/notebooks/mrmr_feature_selection.ipynb
+++ b/docs/notebooks/mrmr_feature_selection.ipynb
@@ -13,17 +13,9 @@
},
{
"cell_type": "code",
- "execution_count": 1,
+ "execution_count": 30,
"metadata": {},
- "outputs": [
- {
- "name": "stderr",
- "output_type": "stream",
- "text": [
- "Using `tqdm.autonotebook.tqdm` in notebook mode. Use `tqdm.tqdm` instead to force console mode (e.g. in jupyter console)\n"
- ]
- }
- ],
+ "outputs": [],
"source": [
"# from IPython.core.display import display, HTML\n",
"# display(HTML(\"\"))\n",
@@ -48,14 +40,14 @@
},
{
"cell_type": "code",
- "execution_count": 2,
+ "execution_count": 31,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
- "Run with ARFS 2.0.5\n"
+ "Run with ARFS 2.2.0\n"
]
}
],
@@ -65,7 +57,7 @@
},
{
"cell_type": "code",
- "execution_count": 3,
+ "execution_count": 32,
"metadata": {},
"outputs": [],
"source": [
@@ -74,16 +66,16 @@
},
{
"cell_type": "code",
- "execution_count": 4,
+ "execution_count": 33,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
- "4"
+ "164"
]
},
- "execution_count": 4,
+ "execution_count": 33,
"metadata": {},
"output_type": "execute_result"
}
@@ -107,7 +99,7 @@
},
{
"cell_type": "code",
- "execution_count": 5,
+ "execution_count": 34,
"metadata": {},
"outputs": [],
"source": [
@@ -118,7 +110,7 @@
},
{
"cell_type": "code",
- "execution_count": 6,
+ "execution_count": 35,
"metadata": {},
"outputs": [
{
@@ -132,7 +124,7 @@
"Name: target, dtype: float64"
]
},
- "execution_count": 6,
+ "execution_count": 35,
"metadata": {},
"output_type": "execute_result"
}
@@ -143,7 +135,7 @@
},
{
"cell_type": "code",
- "execution_count": 7,
+ "execution_count": 36,
"metadata": {},
"outputs": [
{
@@ -170,7 +162,7 @@
"dtype: object"
]
},
- "execution_count": 7,
+ "execution_count": 36,
"metadata": {},
"output_type": "execute_result"
}
@@ -181,7 +173,7 @@
},
{
"cell_type": "code",
- "execution_count": 8,
+ "execution_count": 37,
"metadata": {},
"outputs": [
{
@@ -358,7 +350,7 @@
"4 7.867781 "
]
},
- "execution_count": 8,
+ "execution_count": 37,
"metadata": {},
"output_type": "execute_result"
}
@@ -369,13 +361,13 @@
},
{
"cell_type": "code",
- "execution_count": 9,
+ "execution_count": 38,
"metadata": {},
"outputs": [
{
"data": {
"application/vnd.jupyter.widget-view+json": {
- "model_id": "4abfc0d4fc924b1ab19b7de3657ccc58",
+ "model_id": "f3fba6e1b9bc4b98b5af817e92c0d53e",
"version_major": 2,
"version_minor": 0
},
@@ -386,32 +378,22 @@
"metadata": {},
"output_type": "display_data"
},
- {
- "name": "stderr",
- "output_type": "stream",
- "text": [
- "/home/bsatom/Documents/arfs/src/arfs/association.py:150: RuntimeWarning: Rounded U = 1.0 to 1. This is probably due to floating point precision issues.\n",
- " warnings.warn(\n",
- "/home/bsatom/Documents/arfs/src/arfs/association.py:150: RuntimeWarning: Rounded U = 1.0 to 1. This is probably due to floating point precision issues.\n",
- " warnings.warn(\n"
- ]
- },
{
"data": {
"text/html": [
- "MinRedundancyMaxRelevance(n_features_to_select=5,\n",
- " redundancy_func=functools.partial(<function association_series at 0x7f6055281fc0>, n_jobs=-1, normalize=True),\n",
- " relevance_func=functools.partial(<function f_stat_regression_parallel at 0x7f6055282440>, n_jobs=-1)) In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org. "
+ "MinRedundancyMaxRelevance(n_features_to_select=5, n_jobs=-1,\n",
+ " redundancy_func=functools.partial(<function association_series at 0x000001C620DAA7A0>, n_jobs=-1, normalize=True),\n",
+ " relevance_func=functools.partial(<function f_stat_regression_parallel at 0x000001C620DAAC20>, n_jobs=-1)) In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org. "
],
"text/plain": [
- "MinRedundancyMaxRelevance(n_features_to_select=5,\n",
- " redundancy_func=functools.partial(, n_jobs=-1, normalize=True),\n",
- " relevance_func=functools.partial(, n_jobs=-1))"
+ "MinRedundancyMaxRelevance(n_features_to_select=5, n_jobs=-1,\n",
+ " redundancy_func=functools.partial(, n_jobs=-1, normalize=True),\n",
+ " relevance_func=functools.partial(, n_jobs=-1))"
]
},
- "execution_count": 9,
+ "execution_count": 38,
"metadata": {},
"output_type": "execute_result"
}
@@ -435,7 +417,7 @@
},
{
"cell_type": "code",
- "execution_count": 10,
+ "execution_count": 39,
"metadata": {},
"outputs": [
{
@@ -461,9 +443,9 @@
" \n",
" genuine_num \n",
" LSTAT \n",
+ " RM \n",
" CHAS \n",
" RAD \n",
- " RM \n",
" \n",
" \n",
" \n",
@@ -471,56 +453,56 @@
" 0 \n",
" 7.080332 \n",
" 4.98 \n",
+ " 6.575 \n",
" 0.0 \n",
" 1.0 \n",
- " 6.575 \n",
" \n",
" \n",
" 1 \n",
" 5.245384 \n",
" 9.14 \n",
+ " 6.421 \n",
" 0.0 \n",
" 2.0 \n",
- " 6.421 \n",
" \n",
" \n",
" 2 \n",
" 6.375795 \n",
" 4.03 \n",
+ " 7.185 \n",
" 0.0 \n",
" 2.0 \n",
- " 7.185 \n",
" \n",
" \n",
" 3 \n",
" 6.725118 \n",
" 2.94 \n",
+ " 6.998 \n",
" 0.0 \n",
" 3.0 \n",
- " 6.998 \n",
" \n",
" \n",
" 4 \n",
" 7.867781 \n",
" 5.33 \n",
+ " 7.147 \n",
" 0.0 \n",
" 3.0 \n",
- " 7.147 \n",
" \n",
" \n",
"
\n",
""
],
"text/plain": [
- " genuine_num LSTAT CHAS RAD RM\n",
- "0 7.080332 4.98 0.0 1.0 6.575\n",
- "1 5.245384 9.14 0.0 2.0 6.421\n",
- "2 6.375795 4.03 0.0 2.0 7.185\n",
- "3 6.725118 2.94 0.0 3.0 6.998\n",
- "4 7.867781 5.33 0.0 3.0 7.147"
+ " genuine_num LSTAT RM CHAS RAD\n",
+ "0 7.080332 4.98 6.575 0.0 1.0\n",
+ "1 5.245384 9.14 6.421 0.0 2.0\n",
+ "2 6.375795 4.03 7.185 0.0 2.0\n",
+ "3 6.725118 2.94 6.998 0.0 3.0\n",
+ "4 7.867781 5.33 7.147 0.0 3.0"
]
},
- "execution_count": 10,
+ "execution_count": 39,
"metadata": {},
"output_type": "execute_result"
}
@@ -530,9 +512,78 @@
"X_trans.head()"
]
},
+ {
+ "cell_type": "markdown",
+ "metadata": {},
+ "source": [
+ "Using a single job, avoiding the overhead of starting multiple processes (for moderate size data)"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 40,
+ "metadata": {},
+ "outputs": [
+ {
+ "data": {
+ "application/vnd.jupyter.widget-view+json": {
+ "model_id": "c976973c083a4f05afab5393d4d743e6",
+ "version_major": 2,
+ "version_minor": 0
+ },
+ "text/plain": [
+ " 0%| | 0/5 [00:00, ?it/s]"
+ ]
+ },
+ "metadata": {},
+ "output_type": "display_data"
+ },
+ {
+ "data": {
+ "text/html": [
+ "MinRedundancyMaxRelevance(n_features_to_select=5,\n",
+ " redundancy_func=functools.partial(<function association_series at 0x000001C620DAA7A0>, n_jobs=1, normalize=True),\n",
+ " relevance_func=functools.partial(<function f_stat_regression_parallel at 0x000001C620DAAC20>, n_jobs=1)) In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org. "
+ ],
+ "text/plain": [
+ "MinRedundancyMaxRelevance(n_features_to_select=5,\n",
+ " redundancy_func=functools.partial(, n_jobs=1, normalize=True),\n",
+ " relevance_func=functools.partial(, n_jobs=1))"
+ ]
+ },
+ "execution_count": 40,
+ "metadata": {},
+ "output_type": "execute_result"
+ }
+ ],
+ "source": [
+ "fs_mrmr = MinRedundancyMaxRelevance(\n",
+ " n_features_to_select=5,\n",
+ " relevance_func=None,\n",
+ " redundancy_func=None,\n",
+ " task=\"regression\", # \"classification\",\n",
+ " denominator_func=np.mean,\n",
+ " only_same_domain=False,\n",
+ " return_scores=False,\n",
+ " show_progress=True,\n",
+ " n_jobs=1,\n",
+ ")\n",
+ "\n",
+ "fs_mrmr.fit(X=X, y=y, sample_weight=None)"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": null,
+ "metadata": {},
+ "outputs": [],
+ "source": []
+ },
{
"cell_type": "code",
- "execution_count": 11,
+ "execution_count": 41,
"metadata": {},
"outputs": [
{
@@ -544,7 +595,7 @@
" dtype=object)"
]
},
- "execution_count": 11,
+ "execution_count": 41,
"metadata": {},
"output_type": "execute_result"
}
@@ -555,7 +606,7 @@
},
{
"cell_type": "code",
- "execution_count": 12,
+ "execution_count": 42,
"metadata": {},
"outputs": [
{
@@ -566,7 +617,7 @@
" False])"
]
},
- "execution_count": 12,
+ "execution_count": 42,
"metadata": {},
"output_type": "execute_result"
}
@@ -577,7 +628,7 @@
},
{
"cell_type": "code",
- "execution_count": 13,
+ "execution_count": 43,
"metadata": {},
"outputs": [
{
@@ -586,7 +637,7 @@
"array(['CHAS', 'RM', 'RAD', 'LSTAT', 'genuine_num'], dtype=object)"
]
},
- "execution_count": 13,
+ "execution_count": 43,
"metadata": {},
"output_type": "execute_result"
}
@@ -597,7 +648,7 @@
},
{
"cell_type": "code",
- "execution_count": 14,
+ "execution_count": 44,
"metadata": {},
"outputs": [
{
@@ -640,22 +691,22 @@
" 0.001000 \n",
" \n",
" \n",
+ " RM \n",
+ " 3.249000 \n",
+ " 1.106967 \n",
+ " 0.340710 \n",
+ " \n",
+ " \n",
" CHAS \n",
- " 731.266141 \n",
+ " 1.752553 \n",
" 0.731266 \n",
- " 0.001000 \n",
+ " 0.417258 \n",
" \n",
" \n",
" RAD \n",
- " 3.378313 \n",
+ " 2.070764 \n",
" 0.990593 \n",
- " 0.293221 \n",
- " \n",
- " \n",
- " RM \n",
- " 2.653431 \n",
- " 1.106967 \n",
- " 0.417183 \n",
+ " 0.478371 \n",
" \n",
" \n",
"\n",
@@ -665,12 +716,12 @@
" mrmr relevance redundancy\n",
"genuine_num inf 2.461769 0.000000\n",
"LSTAT 1636.219687 1.636220 0.001000\n",
- "CHAS 731.266141 0.731266 0.001000\n",
- "RAD 3.378313 0.990593 0.293221\n",
- "RM 2.653431 1.106967 0.417183"
+ "RM 3.249000 1.106967 0.340710\n",
+ "CHAS 1.752553 0.731266 0.417258\n",
+ "RAD 2.070764 0.990593 0.478371"
]
},
- "execution_count": 14,
+ "execution_count": 44,
"metadata": {},
"output_type": "execute_result"
}
@@ -689,7 +740,7 @@
},
{
"cell_type": "code",
- "execution_count": 15,
+ "execution_count": 45,
"metadata": {},
"outputs": [],
"source": [
@@ -702,7 +753,7 @@
},
{
"cell_type": "code",
- "execution_count": 16,
+ "execution_count": 46,
"metadata": {},
"outputs": [
{
@@ -741,10 +792,10 @@
"
\n",
" \n",
" 0 \n",
- " 1.0 \n",
+ " 1 \n",
" female \n",
" S \n",
- " Fry \n",
+ " Morty \n",
" 1 \n",
" Mrs \n",
" 29.0000 \n",
@@ -754,10 +805,10 @@
" \n",
" \n",
" 1 \n",
- " 1.0 \n",
+ " 1 \n",
" male \n",
" S \n",
- " Bender \n",
+ " Morty \n",
" 0 \n",
" Master \n",
" 0.9167 \n",
@@ -767,10 +818,10 @@
" \n",
" \n",
" 2 \n",
- " 1.0 \n",
+ " 1 \n",
" female \n",
" S \n",
- " Thanos \n",
+ " Fry \n",
" 0 \n",
" Mrs \n",
" 2.0000 \n",
@@ -780,10 +831,10 @@
" \n",
" \n",
" 3 \n",
- " 1.0 \n",
+ " 1 \n",
" male \n",
" S \n",
- " Morty \n",
+ " Cartman \n",
" 0 \n",
" Mr \n",
" 30.0000 \n",
@@ -793,10 +844,10 @@
" \n",
" \n",
" 4 \n",
- " 1.0 \n",
+ " 1 \n",
" female \n",
" S \n",
- " Morty \n",
+ " Vador \n",
" 0 \n",
" Mrs \n",
" 25.0000 \n",
@@ -810,11 +861,11 @@
],
"text/plain": [
" pclass sex embarked random_cat is_alone title age family_size \\\n",
- "0 1.0 female S Fry 1 Mrs 29.0000 0.0 \n",
- "1 1.0 male S Bender 0 Master 0.9167 3.0 \n",
- "2 1.0 female S Thanos 0 Mrs 2.0000 3.0 \n",
- "3 1.0 male S Morty 0 Mr 30.0000 3.0 \n",
- "4 1.0 female S Morty 0 Mrs 25.0000 3.0 \n",
+ "0 1 female S Morty 1 Mrs 29.0000 0.0 \n",
+ "1 1 male S Morty 0 Master 0.9167 3.0 \n",
+ "2 1 female S Fry 0 Mrs 2.0000 3.0 \n",
+ "3 1 male S Cartman 0 Mr 30.0000 3.0 \n",
+ "4 1 female S Vador 0 Mrs 25.0000 3.0 \n",
"\n",
" fare random_num \n",
"0 211.3375 0.496714 \n",
@@ -824,7 +875,7 @@
"4 151.5500 -0.234153 "
]
},
- "execution_count": 16,
+ "execution_count": 46,
"metadata": {},
"output_type": "execute_result"
}
@@ -835,7 +886,7 @@
},
{
"cell_type": "code",
- "execution_count": 17,
+ "execution_count": 47,
"metadata": {},
"outputs": [
{
@@ -854,7 +905,7 @@
"dtype: object"
]
},
- "execution_count": 17,
+ "execution_count": 47,
"metadata": {},
"output_type": "execute_result"
}
@@ -865,7 +916,7 @@
},
{
"cell_type": "code",
- "execution_count": 18,
+ "execution_count": 48,
"metadata": {},
"outputs": [
{
@@ -886,7 +937,7 @@
"Categories (2, object): ['0', '1']"
]
},
- "execution_count": 18,
+ "execution_count": 48,
"metadata": {},
"output_type": "execute_result"
}
@@ -897,21 +948,13 @@
},
{
"cell_type": "code",
- "execution_count": 19,
+ "execution_count": 49,
"metadata": {},
"outputs": [
- {
- "name": "stderr",
- "output_type": "stream",
- "text": [
- "/home/bsatom/Documents/arfs/src/arfs/association.py:150: RuntimeWarning: Rounded U = 1.0 to 1. This is probably due to floating point precision issues.\n",
- " warnings.warn(\n"
- ]
- },
{
"data": {
"application/vnd.jupyter.widget-view+json": {
- "model_id": "57f07f571dd74d569758463ff2ef66db",
+ "model_id": "e2e5c0dcdeff48d682fe5ae617f67253",
"version_major": 2,
"version_minor": 0
},
@@ -922,35 +965,25 @@
"metadata": {},
"output_type": "display_data"
},
- {
- "name": "stderr",
- "output_type": "stream",
- "text": [
- "/home/bsatom/Documents/arfs/src/arfs/association.py:150: RuntimeWarning: Rounded U = 1.0 to 1. This is probably due to floating point precision issues.\n",
- " warnings.warn(\n",
- "/home/bsatom/Documents/arfs/src/arfs/association.py:150: RuntimeWarning: Rounded U = 1.0 to 1. This is probably due to floating point precision issues.\n",
- " warnings.warn(\n"
- ]
- },
{
"data": {
"text/html": [
- "MinRedundancyMaxRelevance(n_features_to_select=5,\n",
- " redundancy_func=functools.partial(<function association_series at 0x7f6055281fc0>, n_jobs=-1, normalize=True),\n",
- " relevance_func=functools.partial(<function f_stat_classification_parallel at 0x7f6055282680>, n_jobs=-1),\n",
- " task='classification') In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org. MinRedundancyMaxRelevance MinRedundancyMaxRelevance(n_features_to_select=5,\n",
- " redundancy_func=functools.partial(<function association_series at 0x7f6055281fc0>, n_jobs=-1, normalize=True),\n",
- " relevance_func=functools.partial(<function f_stat_classification_parallel at 0x7f6055282680>, n_jobs=-1),\n",
+ "MinRedundancyMaxRelevance(n_features_to_select=5,\n",
+ " redundancy_func=functools.partial(<function association_series at 0x000001C620DAA7A0>, n_jobs=1, normalize=True),\n",
+ " relevance_func=functools.partial(<function f_stat_classification_parallel at 0x000001C620DAAE60>, n_jobs=1),\n",
+ " task='classification') In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org. "
],
"text/plain": [
"MinRedundancyMaxRelevance(n_features_to_select=5,\n",
- " redundancy_func=functools.partial(, n_jobs=-1, normalize=True),\n",
- " relevance_func=functools.partial(, n_jobs=-1),\n",
+ " redundancy_func=functools.partial(, n_jobs=1, normalize=True),\n",
+ " relevance_func=functools.partial(, n_jobs=1),\n",
" task='classification')"
]
},
- "execution_count": 19,
+ "execution_count": 49,
"metadata": {},
"output_type": "execute_result"
}
@@ -965,7 +998,7 @@
" only_same_domain=False,\n",
" return_scores=False,\n",
" show_progress=True,\n",
- " n_jobs=-1,\n",
+ " n_jobs=1,\n",
")\n",
"\n",
"# fs_mrmr.fit(X=X, y=y.astype(str), sample_weight=None)\n",
@@ -974,7 +1007,7 @@
},
{
"cell_type": "code",
- "execution_count": 20,
+ "execution_count": 50,
"metadata": {},
"outputs": [
{
@@ -984,7 +1017,7 @@
" 'age', 'family_size', 'fare', 'random_num', 'target'], dtype=object)"
]
},
- "execution_count": 20,
+ "execution_count": 50,
"metadata": {},
"output_type": "execute_result"
}
@@ -995,17 +1028,17 @@
},
{
"cell_type": "code",
- "execution_count": 21,
+ "execution_count": 51,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
- "array([False, True, False, False, False, True, True, True, True,\n",
+ "array([False, True, False, False, True, True, False, True, True,\n",
" False, False])"
]
},
- "execution_count": 21,
+ "execution_count": 51,
"metadata": {},
"output_type": "execute_result"
}
@@ -1016,16 +1049,16 @@
},
{
"cell_type": "code",
- "execution_count": 22,
+ "execution_count": 52,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
- "array(['sex', 'title', 'age', 'family_size', 'fare'], dtype=object)"
+ "array(['sex', 'is_alone', 'title', 'family_size', 'fare'], dtype=object)"
]
},
- "execution_count": 22,
+ "execution_count": 52,
"metadata": {},
"output_type": "execute_result"
}
@@ -1036,7 +1069,7 @@
},
{
"cell_type": "code",
- "execution_count": 23,
+ "execution_count": 53,
"metadata": {},
"outputs": [
{
@@ -1069,32 +1102,32 @@
" \n",
" sex \n",
" inf \n",
- " 1.740237 \n",
+ " 1.740256 \n",
" 0.000000 \n",
" \n",
" \n",
" fare \n",
- " 1.532597 \n",
+ " 8.129865 \n",
" 1.499352 \n",
- " 0.978309 \n",
+ " 0.184425 \n",
" \n",
" \n",
" title \n",
- " 1.498575 \n",
- " 0.694140 \n",
- " 0.463200 \n",
+ " 1.483999 \n",
+ " 0.694114 \n",
+ " 0.467732 \n",
" \n",
" \n",
" family_size \n",
- " -0.690212 \n",
+ " -1.320894 \n",
" -0.516471 \n",
- " 0.748279 \n",
+ " 0.391001 \n",
" \n",
" \n",
- " age \n",
- " -1.553565 \n",
- " -0.458487 \n",
- " 0.295119 \n",
+ " is_alone \n",
+ " -2.068465 \n",
+ " -0.639219 \n",
+ " 0.309031 \n",
" \n",
" \n",
"\n",
@@ -1102,14 +1135,14 @@
],
"text/plain": [
" mrmr relevance redundancy\n",
- "sex inf 1.740237 0.000000\n",
- "fare 1.532597 1.499352 0.978309\n",
- "title 1.498575 0.694140 0.463200\n",
- "family_size -0.690212 -0.516471 0.748279\n",
- "age -1.553565 -0.458487 0.295119"
+ "sex inf 1.740256 0.000000\n",
+ "fare 8.129865 1.499352 0.184425\n",
+ "title 1.483999 0.694114 0.467732\n",
+ "family_size -1.320894 -0.516471 0.391001\n",
+ "is_alone -2.068465 -0.639219 0.309031"
]
},
- "execution_count": 23,
+ "execution_count": 53,
"metadata": {},
"output_type": "execute_result"
}
@@ -1130,21 +1163,13 @@
},
{
"cell_type": "code",
- "execution_count": 24,
+ "execution_count": 54,
"metadata": {},
"outputs": [
- {
- "name": "stderr",
- "output_type": "stream",
- "text": [
- "/home/bsatom/Documents/arfs/src/arfs/association.py:150: RuntimeWarning: Rounded U = 1.0 to 1. This is probably due to floating point precision issues.\n",
- " warnings.warn(\n"
- ]
- },
{
"data": {
"application/vnd.jupyter.widget-view+json": {
- "model_id": "4560076481cc463ca261ab676c17887a",
+ "model_id": "b5b4c1ac78db415197f6c65f9606633c",
"version_major": 2,
"version_minor": 0
},
@@ -1155,16 +1180,6 @@
"metadata": {},
"output_type": "display_data"
},
- {
- "name": "stderr",
- "output_type": "stream",
- "text": [
- "/home/bsatom/Documents/arfs/src/arfs/association.py:150: RuntimeWarning: Rounded U = 1.0 to 1. This is probably due to floating point precision issues.\n",
- " warnings.warn(\n",
- "/home/bsatom/Documents/arfs/src/arfs/association.py:150: RuntimeWarning: Rounded U = 1.0 to 1. This is probably due to floating point precision issues.\n",
- " warnings.warn(\n"
- ]
- },
{
"data": {
"text/html": [
@@ -1190,7 +1205,7 @@
"
fare \n",
"
title \n",
"
family_size \n",
- "
age \n",
+ "
is_alone \n",
" \n",
" \n",
"
\n",
@@ -1200,7 +1215,7 @@
" 211.3375 \n",
" Mrs \n",
" 0.0 \n",
- " 29.0000 \n",
+ " 1 \n",
" \n",
" \n",
" 1 \n",
@@ -1208,7 +1223,7 @@
" 151.5500 \n",
" Master \n",
" 3.0 \n",
- " 0.9167 \n",
+ " 0 \n",
" \n",
" \n",
" 2 \n",
@@ -1216,7 +1231,7 @@
" 151.5500 \n",
" Mrs \n",
" 3.0 \n",
- " 2.0000 \n",
+ " 0 \n",
" \n",
" \n",
" 3 \n",
@@ -1224,7 +1239,7 @@
" 151.5500 \n",
" Mr \n",
" 3.0 \n",
- " 30.0000 \n",
+ " 0 \n",
" \n",
" \n",
" 4 \n",
@@ -1232,22 +1247,22 @@
" 151.5500 \n",
" Mrs \n",
" 3.0 \n",
- " 25.0000 \n",
+ " 0 \n",
" \n",
" \n",
"\n",
"
"
],
"text/plain": [
- " sex fare title family_size age\n",
- "0 female 211.3375 Mrs 0.0 29.0000\n",
- "1 male 151.5500 Master 3.0 0.9167\n",
- "2 female 151.5500 Mrs 3.0 2.0000\n",
- "3 male 151.5500 Mr 3.0 30.0000\n",
- "4 female 151.5500 Mrs 3.0 25.0000"
+ " sex fare title family_size is_alone\n",
+ "0 female 211.3375 Mrs 0.0 1\n",
+ "1 male 151.5500 Master 3.0 0\n",
+ "2 female 151.5500 Mrs 3.0 0\n",
+ "3 male 151.5500 Mr 3.0 0\n",
+ "4 female 151.5500 Mrs 3.0 0"
]
},
- "execution_count": 24,
+ "execution_count": 54,
"metadata": {},
"output_type": "execute_result"
}
@@ -1275,7 +1290,7 @@
" only_same_domain=False,\n",
" return_scores=False,\n",
" show_progress=True,\n",
- " n_jobs=-1,\n",
+ " n_jobs=1,\n",
")\n",
"\n",
"mrmr_fs_pipeline = Pipeline(\n",
@@ -1294,109 +1309,109 @@
},
{
"cell_type": "code",
- "execution_count": 25,
+ "execution_count": 55,
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"\n",
- "
\n",
+ "\n",
" \n",
" \n",
" \n",
- " predictor \n",
- " missing \n",
- " unique \n",
- " mrmr \n",
+ " predictor \n",
+ " missing \n",
+ " unique \n",
+ " mrmr \n",
" \n",
" \n",
" \n",
" \n",
- " 0 \n",
- " pclass \n",
- " 1 \n",
- " 1 \n",
- " 0 \n",
+ " 0 \n",
+ " pclass \n",
+ " 1 \n",
+ " 1 \n",
+ " 0 \n",
" \n",
" \n",
- " 1 \n",
- " sex \n",
- " 1 \n",
- " 1 \n",
- " 1 \n",
+ " 1 \n",
+ " sex \n",
+ " 1 \n",
+ " 1 \n",
+ " 1 \n",
" \n",
" \n",
- " 2 \n",
- " embarked \n",
- " 1 \n",
- " 1 \n",
- " 0 \n",
+ " 2 \n",
+ " embarked \n",
+ " 1 \n",
+ " 1 \n",
+ " 0 \n",
" \n",
" \n",
- " 3 \n",
- " random_cat \n",
- " 1 \n",
- " 1 \n",
- " 0 \n",
+ " 3 \n",
+ " random_cat \n",
+ " 1 \n",
+ " 1 \n",
+ " 0 \n",
" \n",
" \n",
- " 4 \n",
- " is_alone \n",
- " 1 \n",
- " 1 \n",
- " 0 \n",
+ " 4 \n",
+ " is_alone \n",
+ " 1 \n",
+ " 1 \n",
+ " 1 \n",
" \n",
" \n",
- " 5 \n",
- " title \n",
- " 1 \n",
- " 1 \n",
- " 1 \n",
+ " 5 \n",
+ " title \n",
+ " 1 \n",
+ " 1 \n",
+ " 1 \n",
" \n",
" \n",
- " 6 \n",
- " age \n",
- " 1 \n",
- " 1 \n",
- " 1 \n",
+ " 6 \n",
+ " age \n",
+ " 1 \n",
+ " 1 \n",
+ " 0 \n",
" \n",
" \n",
- " 7 \n",
- " family_size \n",
- " 1 \n",
- " 1 \n",
- " 1 \n",
+ " 7 \n",
+ " family_size \n",
+ " 1 \n",
+ " 1 \n",
+ " 1 \n",
" \n",
" \n",
- " 8 \n",
- " fare \n",
- " 1 \n",
- " 1 \n",
- " 1 \n",
+ " 8 \n",
+ " fare \n",
+ " 1 \n",
+ " 1 \n",
+ " 1 \n",
" \n",
" \n",
- " 9 \n",
- " random_num \n",
- " 1 \n",
- " 1 \n",
- " 0 \n",
+ " 9 \n",
+ " random_num \n",
+ " 1 \n",
+ " 1 \n",
+ " 0 \n",
" \n",
" \n",
"
\n"
],
"text/plain": [
- ""
+ ""
]
},
- "execution_count": 25,
+ "execution_count": 55,
"metadata": {},
"output_type": "execute_result"
}
@@ -1407,16 +1422,16 @@
},
{
"cell_type": "code",
- "execution_count": 26,
+ "execution_count": 56,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
- "array(['sex', 'title', 'age', 'family_size', 'fare'], dtype=object)"
+ "array(['sex', 'is_alone', 'title', 'family_size', 'fare'], dtype=object)"
]
},
- "execution_count": 26,
+ "execution_count": 56,
"metadata": {},
"output_type": "execute_result"
}
@@ -1437,21 +1452,13 @@
},
{
"cell_type": "code",
- "execution_count": 27,
+ "execution_count": 57,
"metadata": {},
"outputs": [
- {
- "name": "stderr",
- "output_type": "stream",
- "text": [
- "/home/bsatom/Documents/arfs/src/arfs/association.py:150: RuntimeWarning: Rounded U = 1.0 to 1. This is probably due to floating point precision issues.\n",
- " warnings.warn(\n"
- ]
- },
{
"data": {
"application/vnd.jupyter.widget-view+json": {
- "model_id": "2ea677aa7ff54dc89b41557690687f73",
+ "model_id": "2e19c29c388547afa819d5240021c903",
"version_major": 2,
"version_minor": 0
},
@@ -1462,20 +1469,6 @@
"metadata": {},
"output_type": "display_data"
},
- {
- "name": "stderr",
- "output_type": "stream",
- "text": [
- "/home/bsatom/Documents/arfs/src/arfs/association.py:150: RuntimeWarning: Rounded U = 1.0 to 1. This is probably due to floating point precision issues.\n",
- " warnings.warn(\n",
- "/home/bsatom/Documents/arfs/src/arfs/association.py:150: RuntimeWarning: Rounded U = 1.0 to 1. This is probably due to floating point precision issues.\n",
- " warnings.warn(\n",
- "/home/bsatom/Documents/arfs/src/arfs/association.py:150: RuntimeWarning: Rounded U = 1.0 to 1. This is probably due to floating point precision issues.\n",
- " warnings.warn(\n",
- "/home/bsatom/Documents/arfs/src/arfs/association.py:150: RuntimeWarning: Rounded U = 1.0 to 1. This is probably due to floating point precision issues.\n",
- " warnings.warn(\n"
- ]
- },
{
"data": {
"text/html": [
@@ -1510,7 +1503,7 @@
" female \n",
" Mrs \n",
" 1 \n",
- " 1.0 \n",
+ " 1 \n",
" S / missing / Q \n",
" \n",
" \n",
@@ -1518,7 +1511,7 @@
" male \n",
" Master \n",
" 0 \n",
- " 1.0 \n",
+ " 1 \n",
" S / missing / Q \n",
" \n",
" \n",
@@ -1526,7 +1519,7 @@
" female \n",
" Mrs \n",
" 0 \n",
- " 1.0 \n",
+ " 1 \n",
" S / missing / Q \n",
" \n",
" \n",
@@ -1534,7 +1527,7 @@
" male \n",
" Mr \n",
" 0 \n",
- " 1.0 \n",
+ " 1 \n",
" S / missing / Q \n",
" \n",
" \n",
@@ -1542,7 +1535,7 @@
" female \n",
" Mrs \n",
" 0 \n",
- " 1.0 \n",
+ " 1 \n",
" S / missing / Q \n",
" \n",
" \n",
@@ -1551,14 +1544,14 @@
],
"text/plain": [
" sex title is_alone pclass embarked\n",
- "0 female Mrs 1 1.0 S / missing / Q\n",
- "1 male Master 0 1.0 S / missing / Q\n",
- "2 female Mrs 0 1.0 S / missing / Q\n",
- "3 male Mr 0 1.0 S / missing / Q\n",
- "4 female Mrs 0 1.0 S / missing / Q"
+ "0 female Mrs 1 1 S / missing / Q\n",
+ "1 male Master 0 1 S / missing / Q\n",
+ "2 female Mrs 0 1 S / missing / Q\n",
+ "3 male Mr 0 1 S / missing / Q\n",
+ "4 female Mrs 0 1 S / missing / Q"
]
},
- "execution_count": 27,
+ "execution_count": 57,
"metadata": {},
"output_type": "execute_result"
}
@@ -1585,7 +1578,7 @@
" only_same_domain=False,\n",
" return_scores=False,\n",
" show_progress=True,\n",
- " n_jobs=-1,\n",
+ " n_jobs=1,\n",
")\n",
"\n",
"mrmr_fs_pipeline = Pipeline(\n",
@@ -1605,124 +1598,124 @@
},
{
"cell_type": "code",
- "execution_count": 28,
+ "execution_count": 58,
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"\n",
- "\n",
+ "\n",
" \n",
" \n",
" \n",
- " predictor \n",
- " missing \n",
- " unique \n",
- " discretizer \n",
- " mrmr \n",
+ " predictor \n",
+ " missing \n",
+ " unique \n",
+ " discretizer \n",
+ " mrmr \n",
" \n",
" \n",
" \n",
" \n",
- " 0 \n",
- " pclass \n",
- " 1 \n",
- " 1 \n",
- " nan \n",
- " 1 \n",
+ " 0 \n",
+ " pclass \n",
+ " 1 \n",
+ " 1 \n",
+ " nan \n",
+ " 1 \n",
" \n",
" \n",
- " 1 \n",
- " sex \n",
- " 1 \n",
- " 1 \n",
- " nan \n",
- " 1 \n",
+ " 1 \n",
+ " sex \n",
+ " 1 \n",
+ " 1 \n",
+ " nan \n",
+ " 1 \n",
" \n",
" \n",
- " 2 \n",
- " embarked \n",
- " 1 \n",
- " 1 \n",
- " nan \n",
- " 1 \n",
+ " 2 \n",
+ " embarked \n",
+ " 1 \n",
+ " 1 \n",
+ " nan \n",
+ " 1 \n",
" \n",
" \n",
- " 3 \n",
- " random_cat \n",
- " 1 \n",
- " 1 \n",
- " nan \n",
- " 0 \n",
+ " 3 \n",
+ " random_cat \n",
+ " 1 \n",
+ " 1 \n",
+ " nan \n",
+ " 0 \n",
" \n",
" \n",
- " 4 \n",
- " is_alone \n",
- " 1 \n",
- " 1 \n",
- " nan \n",
- " 1 \n",
+ " 4 \n",
+ " is_alone \n",
+ " 1 \n",
+ " 1 \n",
+ " nan \n",
+ " 1 \n",
" \n",
" \n",
- " 5 \n",
- " title \n",
- " 1 \n",
- " 1 \n",
- " nan \n",
- " 1 \n",
+ " 5 \n",
+ " title \n",
+ " 1 \n",
+ " 1 \n",
+ " nan \n",
+ " 1 \n",
" \n",
" \n",
- " 6 \n",
- " age \n",
- " 1 \n",
- " 1 \n",
- " nan \n",
- " 0 \n",
+ " 6 \n",
+ " age \n",
+ " 1 \n",
+ " 1 \n",
+ " nan \n",
+ " 0 \n",
" \n",
" \n",
- " 7 \n",
- " family_size \n",
- " 1 \n",
- " 1 \n",
- " nan \n",
- " 0 \n",
+ " 7 \n",
+ " family_size \n",
+ " 1 \n",
+ " 1 \n",
+ " nan \n",
+ " 0 \n",
" \n",
" \n",
- " 8 \n",
- " fare \n",
- " 1 \n",
- " 1 \n",
- " nan \n",
- " 0 \n",
+ " 8 \n",
+ " fare \n",
+ " 1 \n",
+ " 1 \n",
+ " nan \n",
+ " 0 \n",
" \n",
" \n",
- " 9 \n",
- " random_num \n",
- " 1 \n",
- " 1 \n",
- " nan \n",
- " 0 \n",
+ " 9 \n",
+ " random_num \n",
+ " 1 \n",
+ " 1 \n",
+ " nan \n",
+ " 0 \n",
" \n",
" \n",
"
\n"
],
"text/plain": [
- ""
+ ""
]
},
- "execution_count": 28,
+ "execution_count": 58,
"metadata": {},
"output_type": "execute_result"
}
@@ -1748,7 +1741,7 @@
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
- "version": "3.10.12"
+ "version": "3.10.0"
},
"vscode": {
"interpreter": {
diff --git a/docs/notebooks/mrmr_fs_VS_arfs.ipynb b/docs/notebooks/mrmr_fs_VS_arfs.ipynb
index 1bd32c4..efedafd 100644
--- a/docs/notebooks/mrmr_fs_VS_arfs.ipynb
+++ b/docs/notebooks/mrmr_fs_VS_arfs.ipynb
@@ -13,17 +13,9 @@
},
{
"cell_type": "code",
- "execution_count": 1,
+ "execution_count": 22,
"metadata": {},
- "outputs": [
- {
- "name": "stderr",
- "output_type": "stream",
- "text": [
- "Using `tqdm.autonotebook.tqdm` in notebook mode. Use `tqdm.tqdm` instead to force console mode (e.g. in jupyter console)\n"
- ]
- }
- ],
+ "outputs": [],
"source": [
"# from IPython.core.display import display, HTML\n",
"# display(HTML(\"\"))\n",
@@ -56,14 +48,14 @@
},
{
"cell_type": "code",
- "execution_count": 2,
+ "execution_count": 23,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
- "Run with ARFS 2.0.5\n"
+ "Run with ARFS 2.2.0\n"
]
}
],
@@ -73,7 +65,7 @@
},
{
"cell_type": "code",
- "execution_count": 3,
+ "execution_count": 24,
"metadata": {},
"outputs": [],
"source": [
@@ -82,16 +74,16 @@
},
{
"cell_type": "code",
- "execution_count": 4,
+ "execution_count": 25,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
- "4"
+ "12893"
]
},
- "execution_count": 4,
+ "execution_count": 25,
"metadata": {},
"output_type": "execute_result"
}
@@ -113,7 +105,7 @@
},
{
"cell_type": "code",
- "execution_count": 5,
+ "execution_count": 26,
"metadata": {},
"outputs": [],
"source": [
@@ -124,7 +116,7 @@
},
{
"cell_type": "code",
- "execution_count": 6,
+ "execution_count": 27,
"metadata": {},
"outputs": [
{
@@ -138,7 +130,7 @@
"Name: target, dtype: float64"
]
},
- "execution_count": 6,
+ "execution_count": 27,
"metadata": {},
"output_type": "execute_result"
}
@@ -149,7 +141,7 @@
},
{
"cell_type": "code",
- "execution_count": 7,
+ "execution_count": 28,
"metadata": {},
"outputs": [
{
@@ -176,7 +168,7 @@
"dtype: object"
]
},
- "execution_count": 7,
+ "execution_count": 28,
"metadata": {},
"output_type": "execute_result"
}
@@ -187,7 +179,7 @@
},
{
"cell_type": "code",
- "execution_count": 8,
+ "execution_count": 29,
"metadata": {},
"outputs": [
{
@@ -364,7 +356,7 @@
"4 7.867781 "
]
},
- "execution_count": 8,
+ "execution_count": 29,
"metadata": {},
"output_type": "execute_result"
}
@@ -375,13 +367,13 @@
},
{
"cell_type": "code",
- "execution_count": 9,
+ "execution_count": 30,
"metadata": {},
"outputs": [
{
"data": {
"application/vnd.jupyter.widget-view+json": {
- "model_id": "f6189672f38c4c4faec0972721c4cf53",
+ "model_id": "ee5f4ee380b84ead989e2e575219d81d",
"version_major": 2,
"version_minor": 0
},
@@ -392,32 +384,22 @@
"metadata": {},
"output_type": "display_data"
},
- {
- "name": "stderr",
- "output_type": "stream",
- "text": [
- "/home/bsatom/Documents/arfs/src/arfs/association.py:150: RuntimeWarning: Rounded U = 1.0 to 1. This is probably due to floating point precision issues.\n",
- " warnings.warn(\n",
- "/home/bsatom/Documents/arfs/src/arfs/association.py:150: RuntimeWarning: Rounded U = 1.0 to 1. This is probably due to floating point precision issues.\n",
- " warnings.warn(\n"
- ]
- },
{
"data": {
"text/html": [
- "MinRedundancyMaxRelevance(n_features_to_select=10,\n",
- " redundancy_func=functools.partial(<function association_series at 0x7f4f89b21fc0>, n_jobs=-1, normalize=True),\n",
- " relevance_func=functools.partial(<function f_stat_regression_parallel at 0x7f4f89b22440>, n_jobs=-1)) In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org. "
+ "MinRedundancyMaxRelevance(n_features_to_select=10,\n",
+ " redundancy_func=functools.partial(<function association_series at 0x000002D45EFCE7A0>, n_jobs=1, normalize=True),\n",
+ " relevance_func=functools.partial(<function f_stat_regression_parallel at 0x000002D45EFCEC20>, n_jobs=1)) In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org. "
],
"text/plain": [
"MinRedundancyMaxRelevance(n_features_to_select=10,\n",
- " redundancy_func=functools.partial(, n_jobs=-1, normalize=True),\n",
- " relevance_func=functools.partial(, n_jobs=-1))"
+ " redundancy_func=functools.partial(, n_jobs=1, normalize=True),\n",
+ " relevance_func=functools.partial(, n_jobs=1))"
]
},
- "execution_count": 9,
+ "execution_count": 30,
"metadata": {},
"output_type": "execute_result"
}
@@ -434,7 +416,7 @@
" only_same_domain=False,\n",
" return_scores=False,\n",
" show_progress=True,\n",
- " n_jobs=-1,\n",
+ " n_jobs=1,\n",
")\n",
"\n",
"# fs_mrmr.fit(X=X, y=y.astype(str), sample_weight=None)\n",
@@ -443,7 +425,7 @@
},
{
"cell_type": "code",
- "execution_count": 10,
+ "execution_count": 31,
"metadata": {},
"outputs": [
{
@@ -469,14 +451,14 @@
" \n",
" genuine_num \n",
" LSTAT \n",
+ " RM \n",
" CHAS \n",
" RAD \n",
- " RM \n",
" PTRATIO \n",
" INDUS \n",
" TAX \n",
" NOX \n",
- " AGE \n",
+ " CRIM \n",
" \n",
" \n",
" \n",
@@ -484,81 +466,81 @@
" 0 \n",
" 7.080332 \n",
" 4.98 \n",
+ " 6.575 \n",
" 0.0 \n",
" 1.0 \n",
- " 6.575 \n",
" 15.3 \n",
" 2.31 \n",
" 296.0 \n",
" 0.538 \n",
- " 65.2 \n",
+ " 0.00632 \n",
" \n",
" \n",
" 1 \n",
" 5.245384 \n",
" 9.14 \n",
+ " 6.421 \n",
" 0.0 \n",
" 2.0 \n",
- " 6.421 \n",
" 17.8 \n",
" 7.07 \n",
" 242.0 \n",
" 0.469 \n",
- " 78.9 \n",
+ " 0.02731 \n",
" \n",
" \n",
" 2 \n",
" 6.375795 \n",
" 4.03 \n",
+ " 7.185 \n",
" 0.0 \n",
" 2.0 \n",
- " 7.185 \n",
" 17.8 \n",
" 7.07 \n",
" 242.0 \n",
" 0.469 \n",
- " 61.1 \n",
+ " 0.02729 \n",
" \n",
" \n",
" 3 \n",
" 6.725118 \n",
" 2.94 \n",
+ " 6.998 \n",
" 0.0 \n",
" 3.0 \n",
- " 6.998 \n",
" 18.7 \n",
" 2.18 \n",
" 222.0 \n",
" 0.458 \n",
- " 45.8 \n",
+ " 0.03237 \n",
" \n",
" \n",
" 4 \n",
" 7.867781 \n",
" 5.33 \n",
+ " 7.147 \n",
" 0.0 \n",
" 3.0 \n",
- " 7.147 \n",
" 18.7 \n",
" 2.18 \n",
" 222.0 \n",
" 0.458 \n",
- " 54.2 \n",
+ " 0.06905 \n",
" \n",
" \n",
"
\n",
""
],
"text/plain": [
- " genuine_num LSTAT CHAS RAD RM PTRATIO INDUS TAX NOX AGE\n",
- "0 7.080332 4.98 0.0 1.0 6.575 15.3 2.31 296.0 0.538 65.2\n",
- "1 5.245384 9.14 0.0 2.0 6.421 17.8 7.07 242.0 0.469 78.9\n",
- "2 6.375795 4.03 0.0 2.0 7.185 17.8 7.07 242.0 0.469 61.1\n",
- "3 6.725118 2.94 0.0 3.0 6.998 18.7 2.18 222.0 0.458 45.8\n",
- "4 7.867781 5.33 0.0 3.0 7.147 18.7 2.18 222.0 0.458 54.2"
+ " genuine_num LSTAT RM CHAS RAD PTRATIO INDUS TAX NOX CRIM\n",
+ "0 7.080332 4.98 6.575 0.0 1.0 15.3 2.31 296.0 0.538 0.00632\n",
+ "1 5.245384 9.14 6.421 0.0 2.0 17.8 7.07 242.0 0.469 0.02731\n",
+ "2 6.375795 4.03 7.185 0.0 2.0 17.8 7.07 242.0 0.469 0.02729\n",
+ "3 6.725118 2.94 6.998 0.0 3.0 18.7 2.18 222.0 0.458 0.03237\n",
+ "4 7.867781 5.33 7.147 0.0 3.0 18.7 2.18 222.0 0.458 0.06905"
]
},
- "execution_count": 10,
+ "execution_count": 31,
"metadata": {},
"output_type": "execute_result"
}
@@ -570,7 +552,7 @@
},
{
"cell_type": "code",
- "execution_count": 11,
+ "execution_count": 32,
"metadata": {},
"outputs": [
{
@@ -582,7 +564,7 @@
" dtype=object)"
]
},
- "execution_count": 11,
+ "execution_count": 32,
"metadata": {},
"output_type": "execute_result"
}
@@ -593,18 +575,18 @@
},
{
"cell_type": "code",
- "execution_count": 12,
+ "execution_count": 33,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
- "array([False, False, True, True, True, True, True, False, True,\n",
+ "array([ True, False, True, True, True, True, False, False, True,\n",
" True, True, False, True, False, False, False, False, True,\n",
" False])"
]
},
- "execution_count": 12,
+ "execution_count": 33,
"metadata": {},
"output_type": "execute_result"
}
@@ -615,17 +597,17 @@
},
{
"cell_type": "code",
- "execution_count": 13,
+ "execution_count": 34,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
- "array(['INDUS', 'CHAS', 'NOX', 'RM', 'AGE', 'RAD', 'TAX', 'PTRATIO',\n",
+ "array(['CRIM', 'INDUS', 'CHAS', 'NOX', 'RM', 'RAD', 'TAX', 'PTRATIO',\n",
" 'LSTAT', 'genuine_num'], dtype=object)"
]
},
- "execution_count": 13,
+ "execution_count": 34,
"metadata": {},
"output_type": "execute_result"
}
@@ -636,7 +618,7 @@
},
{
"cell_type": "code",
- "execution_count": 14,
+ "execution_count": 35,
"metadata": {},
"outputs": [
{
@@ -679,52 +661,52 @@
" 0.001000 \n",
" \n",
" \n",
+ " RM \n",
+ " 3.249000 \n",
+ " 1.106967 \n",
+ " 0.340710 \n",
+ " \n",
+ " \n",
" CHAS \n",
- " 731.266141 \n",
+ " 1.752553 \n",
" 0.731266 \n",
- " 0.001000 \n",
+ " 0.417258 \n",
" \n",
" \n",
" RAD \n",
- " 3.378313 \n",
+ " 2.070764 \n",
" 0.990593 \n",
- " 0.293221 \n",
- " \n",
- " \n",
- " RM \n",
- " 2.653431 \n",
- " 1.106967 \n",
- " 0.417183 \n",
+ " 0.478371 \n",
" \n",
" \n",
" PTRATIO \n",
- " -0.215423 \n",
+ " -0.326440 \n",
" -0.103248 \n",
- " 0.479280 \n",
+ " 0.316285 \n",
" \n",
" \n",
" INDUS \n",
- " -0.462910 \n",
+ " -0.510767 \n",
" -0.189508 \n",
- " 0.409384 \n",
+ " 0.371026 \n",
" \n",
" \n",
" TAX \n",
- " -0.497833 \n",
+ " -0.503383 \n",
" -0.239237 \n",
- " 0.480558 \n",
+ " 0.475259 \n",
" \n",
" \n",
" NOX \n",
- " -0.655342 \n",
+ " -0.721365 \n",
" -0.358202 \n",
- " 0.546589 \n",
+ " 0.496562 \n",
" \n",
" \n",
- " AGE \n",
- " -0.873075 \n",
- " -0.476940 \n",
- " 0.546275 \n",
+ " CRIM \n",
+ " -0.839506 \n",
+ " -0.452434 \n",
+ " 0.538929 \n",
" \n",
" \n",
"
\n",
@@ -734,17 +716,17 @@
" mrmr relevance redundancy\n",
"genuine_num inf 2.461769 0.000000\n",
"LSTAT 1636.219687 1.636220 0.001000\n",
- "CHAS 731.266141 0.731266 0.001000\n",
- "RAD 3.378313 0.990593 0.293221\n",
- "RM 2.653431 1.106967 0.417183\n",
- "PTRATIO -0.215423 -0.103248 0.479280\n",
- "INDUS -0.462910 -0.189508 0.409384\n",
- "TAX -0.497833 -0.239237 0.480558\n",
- "NOX -0.655342 -0.358202 0.546589\n",
- "AGE -0.873075 -0.476940 0.546275"
+ "RM 3.249000 1.106967 0.340710\n",
+ "CHAS 1.752553 0.731266 0.417258\n",
+ "RAD 2.070764 0.990593 0.478371\n",
+ "PTRATIO -0.326440 -0.103248 0.316285\n",
+ "INDUS -0.510767 -0.189508 0.371026\n",
+ "TAX -0.503383 -0.239237 0.475259\n",
+ "NOX -0.721365 -0.358202 0.496562\n",
+ "CRIM -0.839506 -0.452434 0.538929"
]
},
- "execution_count": 14,
+ "execution_count": 35,
"metadata": {},
"output_type": "execute_result"
}
@@ -765,7 +747,7 @@
},
{
"cell_type": "code",
- "execution_count": 15,
+ "execution_count": 36,
"metadata": {},
"outputs": [],
"source": [
@@ -776,13 +758,13 @@
},
{
"cell_type": "code",
- "execution_count": 16,
+ "execution_count": 37,
"metadata": {},
"outputs": [
{
"data": {
"application/vnd.jupyter.widget-view+json": {
- "model_id": "140eff2aa1df4ac0a44c39bf9595cf12",
+ "model_id": "05980d7df5fb42ec9ff93ae926dc1cf5",
"version_major": 2,
"version_minor": 0
},
@@ -799,12 +781,12 @@
"text": [
"The selected features: ['CRIM' 'NOX' 'RM' 'AGE' 'DIS' 'TAX' 'PTRATIO' 'LSTAT' 'genuine_num']\n",
"The agnostic ranking: [2 1 1 1 2 2 2 2 1 2 2 1 2 1 1 1 1 2]\n",
- "The naive ranking: ['LSTAT', 'RM', 'genuine_num', 'PTRATIO', 'DIS', 'CRIM', 'NOX', 'AGE', 'TAX', 'B', 'random_num1', 'INDUS', 'random_cat', 'random_cat_2', 'RAD', 'ZN', 'random_num2', 'CHAS']\n"
+ "The naive ranking: ['LSTAT', 'RM', 'genuine_num', 'PTRATIO', 'DIS', 'NOX', 'CRIM', 'AGE', 'TAX', 'B', 'random_num1', 'INDUS', 'random_cat', 'random_cat_2', 'RAD', 'ZN', 'random_num2', 'CHAS']\n"
]
},
{
"data": {
- "image/png": "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",
+ "image/png": "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",
"text/plain": [
"
"
]
@@ -816,8 +798,8 @@
"name": "stdout",
"output_type": "stream",
"text": [
- "CPU times: user 45.2 s, sys: 850 ms, total: 46.1 s\n",
- "Wall time: 16.5 s\n"
+ "CPU times: total: 39.2 s\n",
+ "Wall time: 7.87 s\n"
]
}
],
@@ -847,13 +829,13 @@
},
{
"cell_type": "code",
- "execution_count": 17,
+ "execution_count": 38,
"metadata": {},
"outputs": [
{
"data": {
"application/vnd.jupyter.widget-view+json": {
- "model_id": "8a2a9912551a4cb68e90f1d3fc514b9a",
+ "model_id": "ad08f38cc7a34b74a4cadf8ff8256fe3",
"version_major": 2,
"version_minor": 0
},
@@ -882,18 +864,18 @@
},
{
"cell_type": "code",
- "execution_count": 18,
+ "execution_count": 39,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
- "array(['INDUS', 'CHAS', 'RM', 'RAD', 'PTRATIO', 'random_num1',\n",
- " 'random_num2', 'random_cat', 'random_cat_2', 'genuine_num'],\n",
- " dtype=object)"
+ "array(['CRIM', 'ZN', 'INDUS', 'CHAS', 'NOX', 'RM', 'AGE', 'DIS', 'RAD',\n",
+ " 'TAX', 'PTRATIO', 'B', 'LSTAT', 'random_num1', 'random_num2',\n",
+ " 'random_cat', 'random_cat_2', 'genuine_num'], dtype=object)"
]
},
- "execution_count": 18,
+ "execution_count": 39,
"metadata": {},
"output_type": "execute_result"
}
@@ -904,12 +886,12 @@
},
{
"cell_type": "code",
- "execution_count": 19,
+ "execution_count": 40,
"metadata": {},
"outputs": [
{
"data": {
- "image/png": "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",
+ "image/png": "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",
"text/plain": [
""
]
@@ -924,14 +906,14 @@
},
{
"cell_type": "code",
- "execution_count": 20,
+ "execution_count": 41,
"metadata": {},
"outputs": [
{
"data": {
- "image/png": "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",
+ "image/png": "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",
"text/plain": [
- ""
+ ""
]
},
"metadata": {},
@@ -948,188 +930,184 @@
},
{
"cell_type": "code",
- "execution_count": 21,
+ "execution_count": 42,
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"\n",
- "\n",
+ "\n",
" \n",
" \n",
" \n",
- " predictor \n",
- " missing \n",
- " unique \n",
- " collinearity \n",
- " arfs \n",
+ " predictor \n",
+ " missing \n",
+ " unique \n",
+ " collinearity \n",
+ " arfs \n",
" \n",
" \n",
" \n",
" \n",
- " 0 \n",
- " CRIM \n",
- " 1 \n",
- " 1 \n",
- " 0 \n",
- " nan \n",
+ " 0 \n",
+ " CRIM \n",
+ " 1 \n",
+ " 1 \n",
+ " 1 \n",
+ " 1 \n",
" \n",
" \n",
- " 1 \n",
- " ZN \n",
- " 1 \n",
- " 1 \n",
- " 0 \n",
- " nan \n",
+ " 1 \n",
+ " ZN \n",
+ " 1 \n",
+ " 1 \n",
+ " 1 \n",
+ " 0 \n",
" \n",
" \n",
- " 2 \n",
- " INDUS \n",
- " 1 \n",
- " 1 \n",
- " 1 \n",
- " 1 \n",
+ " 2 \n",
+ " INDUS \n",
+ " 1 \n",
+ " 1 \n",
+ " 1 \n",
+ " 0 \n",
" \n",
" \n",
- " 3 \n",
- " CHAS \n",
- " 1 \n",
- " 1 \n",
- " 1 \n",
- " 0 \n",
+ " 3 \n",
+ " CHAS \n",
+ " 1 \n",
+ " 1 \n",
+ " 1 \n",
+ " 0 \n",
" \n",
" \n",
- " 4 \n",
- " NOX \n",
- " 1 \n",
- " 1 \n",
- " 0 \n",
- " nan \n",
+ " 4 \n",
+ " NOX \n",
+ " 1 \n",
+ " 1 \n",
+ " 1 \n",
+ " 1 \n",
" \n",
" \n",
- " 5 \n",
- " RM \n",
- " 1 \n",
- " 1 \n",
- " 1 \n",
- " 1 \n",
+ " 5 \n",
+ " RM \n",
+ " 1 \n",
+ " 1 \n",
+ " 1 \n",
+ " 1 \n",
" \n",
" \n",
- " 6 \n",
- " AGE \n",
- " 1 \n",
- " 1 \n",
- " 0 \n",
- " nan \n",
+ " 6 \n",
+ " AGE \n",
+ " 1 \n",
+ " 1 \n",
+ " 1 \n",
+ " 1 \n",
" \n",
" \n",
- " 7 \n",
- " DIS \n",
- " 1 \n",
- " 1 \n",
- " 0 \n",
- " nan \n",
+ " 7 \n",
+ " DIS \n",
+ " 1 \n",
+ " 1 \n",
+ " 1 \n",
+ " 1 \n",
" \n",
" \n",
- " 8 \n",
- " RAD \n",
- " 1 \n",
- " 1 \n",
- " 1 \n",
- " 1 \n",
+ " 8 \n",
+ " RAD \n",
+ " 1 \n",
+ " 1 \n",
+ " 1 \n",
+ " 0 \n",
" \n",
" \n",
- " 9 \n",
- " TAX \n",
- " 1 \n",
- " 1 \n",
- " 0 \n",
- " nan \n",
+ " 9 \n",
+ " TAX \n",
+ " 1 \n",
+ " 1 \n",
+ " 1 \n",
+ " 1 \n",
" \n",
" \n",
- " 10 \n",
- " PTRATIO \n",
- " 1 \n",
- " 1 \n",
- " 1 \n",
- " 1 \n",
+ " 10 \n",
+ " PTRATIO \n",
+ " 1 \n",
+ " 1 \n",
+ " 1 \n",
+ " 1 \n",
" \n",
" \n",
- " 11 \n",
- " B \n",
- " 1 \n",
- " 1 \n",
- " 0 \n",
- " nan \n",
+ " 11 \n",
+ " B \n",
+ " 1 \n",
+ " 1 \n",
+ " 1 \n",
+ " 0 \n",
" \n",
" \n",
- " 12 \n",
- " LSTAT \n",
- " 1 \n",
- " 1 \n",
- " 0 \n",
- " nan \n",
+ " 12 \n",
+ " LSTAT \n",
+ " 1 \n",
+ " 1 \n",
+ " 1 \n",
+ " 1 \n",
" \n",
" \n",
- " 13 \n",
- " random_num1 \n",
- " 1 \n",
- " 1 \n",
- " 1 \n",
- " 0 \n",
+ " 13 \n",
+ " random_num1 \n",
+ " 1 \n",
+ " 1 \n",
+ " 1 \n",
+ " 0 \n",
" \n",
" \n",
- " 14 \n",
- " random_num2 \n",
- " 1 \n",
- " 1 \n",
- " 1 \n",
- " 0 \n",
+ " 14 \n",
+ " random_num2 \n",
+ " 1 \n",
+ " 1 \n",
+ " 1 \n",
+ " 0 \n",
" \n",
" \n",
- " 15 \n",
- " random_cat \n",
- " 1 \n",
- " 1 \n",
- " 1 \n",
- " 0 \n",
+ " 15 \n",
+ " random_cat \n",
+ " 1 \n",
+ " 1 \n",
+ " 1 \n",
+ " 0 \n",
" \n",
" \n",
- " 16 \n",
- " random_cat_2 \n",
- " 1 \n",
- " 1 \n",
- " 1 \n",
- " 0 \n",
+ " 16 \n",
+ " random_cat_2 \n",
+ " 1 \n",
+ " 1 \n",
+ " 1 \n",
+ " 0 \n",
" \n",
" \n",
- " 17 \n",
- " genuine_num \n",
- " 1 \n",
- " 1 \n",
- " 1 \n",
- " 1 \n",
+ " 17 \n",
+ " genuine_num \n",
+ " 1 \n",
+ " 1 \n",
+ " 1 \n",
+ " 1 \n",
" \n",
" \n",
"
\n"
],
"text/plain": [
- ""
+ ""
]
},
- "execution_count": 21,
+ "execution_count": 42,
"metadata": {},
"output_type": "execute_result"
}
@@ -1155,7 +1133,7 @@
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
- "version": "3.10.12"
+ "version": "3.10.0"
},
"vscode": {
"interpreter": {
diff --git a/src/arfs/__init__.py b/src/arfs/__init__.py
index 74ee4c6..c782e21 100644
--- a/src/arfs/__init__.py
+++ b/src/arfs/__init__.py
@@ -1,4 +1,4 @@
"""init module, providing information about the arfs package
"""
-__version__ = "2.1.4"
+__version__ = "2.2.0"
diff --git a/src/arfs/association.py b/src/arfs/association.py
index 86d7b8b..9e966bb 100644
--- a/src/arfs/association.py
+++ b/src/arfs/association.py
@@ -20,6 +20,7 @@
from mpl_toolkits.axes_grid1 import make_axes_locatable
from sklearn.utils import as_float_array, safe_sqr, safe_mask
+
from multiprocessing import cpu_count
from itertools import combinations, permutations, product
from pandas.api.types import is_numeric_dtype
@@ -54,30 +55,72 @@
# - the function looping over a chunk of combinations
# - the parallelization (sending different chunks to different cores and applying the latter function)
+
+def symmetric_function(func):
+ func.is_symmetric = True
+ return func
+
+
+def asymmetric_function(func):
+ func.is_symmetric = False
+ return func
+
+
+def create_col_combinations(func, selected_cols):
+ """
+ Create column combinations or permutations based on the symmetry of the function.
+
+ This function checks if `func` is symmetric. If it is, it creates combinations of `selected_cols`;
+ otherwise, it creates permutations.
+
+ Parameters
+ ----------
+ func : callable
+ The function to check for symmetry. Should be decorated with `@symmetric_function`.
+ selected_cols : list
+ The columns to be combined or permuted.
+
+ Returns
+ -------
+ list of tuples
+ A list of tuples representing column combinations or permutations.
+ If `func` is symmetric, combinations of `selected_cols` are returned;
+ otherwise, permutations are returned.
+ """
+
+ if getattr(func, "is_symmetric", False):
+ # If the function is symmetric, use combinations
+ return list(combinations(selected_cols, 2)) if selected_cols else []
+ else:
+ # If the function is not symmetric, use permutations
+ return list(permutations(selected_cols, 2)) if selected_cols else []
+
+
##################
# CAT-CAT
##################
def weighted_conditional_entropy(x, y, sample_weight=None):
- """weighted_conditional_entropy computes the weighted conditional entropy between two
- categorical predictors.
+ """
+ Computes the weighted conditional entropy between two categorical predictors.
Parameters
----------
x : pd.Series of shape (n_samples,)
- The predictor vector
+ The predictor vector.
y : pd.Series of shape (n_samples,)
- The target vector
+ The target vector.
sample_weight : array-like of shape (n_samples,), optional
- The weight vector, by default None
+ The weight vector, by default None.
Returns
-------
float
- weighted conditional entropy
+ Weighted conditional entropy.
"""
+ # Handle sample_weight
if sample_weight is None:
sample_weight = np.ones(len(x))
elif np.count_nonzero(sample_weight) == 0:
@@ -85,81 +128,114 @@ def weighted_conditional_entropy(x, y, sample_weight=None):
"All elements in sample_weight are zero. Cannot divide by zero."
)
- df = pd.DataFrame({"x": x, "y": y, "sample_weight": sample_weight})
- tot_weight = df["sample_weight"].sum()
- y_counter = df[["y", "sample_weight"]].groupby("y").sum().to_dict()
- y_counter = y_counter["sample_weight"]
- xy_counter = df[["x", "y", "sample_weight"]].groupby(["x", "y"]).sum().to_dict()
- xy_counter = xy_counter["sample_weight"]
+ # Integer encoding for categorical data
+ y_encoded, _ = pd.factorize(y)
+ x_encoded, _ = pd.factorize(x)
+
+ # Total weight
+ tot_weight = np.sum(sample_weight)
+ if tot_weight == 0:
+ return 0
+
+ # Grouped weights for y and (x, y)
+ y_weights = np.bincount(
+ y_encoded, weights=sample_weight, minlength=len(np.unique(y_encoded))
+ )
+ xy_weights = {
+ level: np.bincount(
+ y_encoded[x_encoded == level],
+ weights=sample_weight[x_encoded == level],
+ minlength=len(np.unique(y_encoded)),
+ )
+ for level in np.unique(x_encoded)
+ }
+
+ # Conditional entropy calculation
h_xy = 0.0
- for xy in xy_counter.keys():
- p_xy = xy_counter[xy] / tot_weight if tot_weight != 0 else 0
- p_y = y_counter[xy[1]] / tot_weight if tot_weight != 0 else 0
- if p_xy != 0:
- h_xy += p_xy * math.log(p_y / p_xy, math.e)
+ for level in xy_weights:
+ for y_index, xy_weight in enumerate(xy_weights[level]):
+ p_xy = xy_weight / tot_weight
+ p_y = y_weights[y_index] / tot_weight
+
+ if p_xy != 0:
+ h_xy += p_xy * math.log(p_y / p_xy, math.e)
+
return h_xy
+@asymmetric_function
def weighted_theils_u(x, y, sample_weight=None, as_frame=False):
- """weighted_theils_u computes the weighted Theil's U statistic between two
- categorical predictors.
+ """
+ Computes the weighted Theil's U statistic between two categorical predictors.
Parameters
----------
x : pd.Series of shape (n_samples,)
- The predictor vector
+ The predictor vector.
y : pd.Series of shape (n_samples,)
- The target vector
+ The target vector.
sample_weight : array-like of shape (n_samples,), optional
- The weight vector, by default None
- as_frame: bool
- return output as a dataframe or a float
+ The weight vector, by default None.
+ as_frame : bool
+ Return output as a dataframe or a float.
Returns
-------
- pd.DataFrame
- predictor names and value of the Theil's U statistic
+ pd.DataFrame or float
+ Predictor names and value of the Theil's U statistic.
"""
if sample_weight is None:
sample_weight = np.ones(len(x))
- df = pd.DataFrame({"x": x, "y": y, "sample_weight": sample_weight})
- tot_weight = df["sample_weight"].sum()
- y_counter = df[["y", "sample_weight"]].groupby("y").sum().to_dict()
- y_counter = y_counter["sample_weight"]
- x_counter = df[["x", "sample_weight"]].groupby("x").sum().to_dict()
- x_counter = x_counter["sample_weight"]
- p_x = list(map(lambda n: n / tot_weight, x_counter.values()))
+ tot_weight = np.sum(sample_weight)
+
+ # Integer encoding
+ y_encoded, y_unique = pd.factorize(y)
+ x_encoded, x_unique = pd.factorize(x)
+
+ # Extend bincount to cover all categories
+ y_weights = np.bincount(y_encoded, weights=sample_weight, minlength=len(y_unique))
+ x_weights = np.bincount(x_encoded, weights=sample_weight, minlength=len(x_unique))
+
+ # Entropy calculations
+ p_x = x_weights / tot_weight
h_x = ss.entropy(p_x)
- xy_counter = df[["x", "y", "sample_weight"]].groupby(["x", "y"]).sum().to_dict()
- xy_counter = xy_counter["sample_weight"]
+
h_xy = 0.0
- for xy in xy_counter.keys():
- p_xy = xy_counter[xy] / tot_weight if tot_weight != 0 else 0
- p_y = y_counter[xy[1]] / tot_weight if tot_weight != 0 else 0
- if p_xy != 0:
- h_xy += p_xy * math.log(p_y / p_xy, math.e)
+ for unique_x in np.unique(x_encoded):
+ x_mask = x_encoded == unique_x
+ y_sub_weights = np.bincount(
+ y_encoded[x_mask], weights=sample_weight[x_mask], minlength=len(y_unique)
+ )
+ p_xy = y_sub_weights / tot_weight
+ p_y = y_weights / tot_weight
+ # Avoid division by zero in log calculation
+ valid_mask = (p_xy != 0) & (p_y != 0)
+ h_xy += np.sum(p_xy[valid_mask] * np.log(p_y[valid_mask] / p_xy[valid_mask]))
if h_x == 0:
return 1.0
- else:
- u = (h_x - h_xy) / h_x
- if abs(u) < _PRECISION or abs(u - 1.0) < _PRECISION:
- rounded_u = round(u)
- warnings.warn(
- f"Rounded U = {u} to {rounded_u}. This is probably due to floating point precision issues.",
- RuntimeWarning,
- )
- u = rounded_u
+ u = (h_x - h_xy) / h_x
+
+ # Check for floating point precision issues
+ if abs(u) < _PRECISION or abs(u - 1.0) < _PRECISION:
+ rounded_u = round(u)
+ warnings.warn(
+ f"Rounded U = {u} to {rounded_u}. This is probably due to floating point precision issues.",
+ RuntimeWarning,
+ )
+ u = rounded_u
+
+ # Return as DataFrame or float
if as_frame:
- return pd.DataFrame({"row": x.name, "col": y.name, "val": u}, index=[0])
+ return pd.DataFrame({"row": [x.name], "col": [y.name], "val": [u]})
else:
return u
-def theils_u_matrix(X, sample_weight=None, n_jobs=-1, handle_na="drop"):
+def theils_u_matrix(X, sample_weight=None, n_jobs=1, handle_na="drop"):
"""theils_u_matrix theils_u_matrix computes the weighted Theil's U statistic for
categorical-categorical association. This is an asymmetric coefficient: U(x,y) != U(y,x)
U(x, y) means the uncertainty of x given y: value is on the range of [0,1] -
@@ -213,14 +289,14 @@ def theils_u_matrix(X, sample_weight=None, n_jobs=-1, handle_na="drop"):
df=X,
comb_list=comb_list,
sample_weight=sample_weight,
- n_jobs=-1,
+ n_jobs=n_jobs,
)
return lst
else:
return None
-def theils_u_series(X, target, sample_weight=None, n_jobs=-1, handle_na="drop"):
+def theils_u_series(X, target, sample_weight=None, n_jobs=1, handle_na="drop"):
"""theils_u_series computes the weighted Theil's U statistic for
categorical-categorical association. This is an asymmetric coefficient: U(x,y) != U(y,x)
U(x, y) means the uncertainty of x given y: value is on the range of [0,1] -
@@ -280,29 +356,38 @@ def theils_u_series(X, target, sample_weight=None, n_jobs=-1, handle_na="drop"):
return None
+@symmetric_function
def cramer_v(x, y, sample_weight=None, as_frame=False):
- """cramer_v computes the weighted V statistic of two
- categorical predictors.
+ """
+ Computes the weighted V statistic of two categorical predictors.
Parameters
----------
x : pd.Series of shape (n_samples,)
- The predictor vector, the first categorical predictor
+ The first categorical predictor.
y : pd.Series of shape (n_samples,)
- second categorical predictor, order doesn't matter, symmetrical association
+ The second categorical predictor, order doesn't matter, symmetrical association.
sample_weight : array-like of shape (n_samples,), optional
- The weight vector, by default None
- as_frame: bool
- return output as a dataframe or a float
+ The weight vector, by default None.
+ as_frame : bool
+ Return output as a DataFrame or a float.
Returns
-------
- pd.DataFrame
- single row dataframe with the predictor names and the statistic value
+ pd.DataFrame or float
+ Single row DataFrame with the predictor names and the statistic value, or the statistic as a float.
"""
- tot_weight = sample_weight.sum()
+
+ if sample_weight is None:
+ sample_weight = np.ones(len(x))
+ else:
+ sample_weight = np.asarray(sample_weight)
+ if sample_weight.sum() == 0:
+ raise ValueError("Sum of sample weights cannot be zero.")
+
weighted_tab = pd.crosstab(x, y, sample_weight, aggfunc=sum).fillna(0)
- chi2 = ss.chi2_contingency(weighted_tab)[0]
+ chi2 = ss.chi2_contingency(weighted_tab, correction=False)[0]
+ tot_weight = sample_weight.sum()
phi2 = chi2 / tot_weight
r, k = weighted_tab.shape
phi2corr = max(0, phi2 - ((k - 1) * (r - 1)) / (tot_weight - 1))
@@ -311,8 +396,8 @@ def cramer_v(x, y, sample_weight=None, as_frame=False):
v = np.sqrt(phi2corr / min((kcorr - 1), (rcorr - 1)))
if as_frame:
- x_name = x.name if isinstance(x, pd.Series) else "var"
- y_name = y.name if isinstance(y, pd.Series) else "target"
+ x_name = x.name if x.name else "var1"
+ y_name = y.name if y.name else "var2"
return pd.DataFrame(
{"row": [x_name, y_name], "col": [y_name, x_name], "val": [v, v]}
)
@@ -320,7 +405,7 @@ def cramer_v(x, y, sample_weight=None, as_frame=False):
return v
-def cramer_v_matrix(X, sample_weight=None, n_jobs=-1, handle_na="drop"):
+def cramer_v_matrix(X, sample_weight=None, n_jobs=1, handle_na="drop"):
"""cramer_v_matrix computes the weighted Cramer's V statistic for
categorical-categorical association. This is a symmetric coefficient: V(x,y) = V(y,x)
@@ -336,7 +421,7 @@ def cramer_v_matrix(X, sample_weight=None, n_jobs=-1, handle_na="drop"):
sample_weight : array-like of shape (n_samples,), optional
The weight vector, by default None
n_jobs : int, optional
- the number of cores to use for the computation, by default -1
+ the number of cores to use for the computation, by default 1
handle_na : str, optional
either drop rows with na, fill na with 0 or do nothing, by default "drop"
@@ -375,7 +460,7 @@ def cramer_v_matrix(X, sample_weight=None, n_jobs=-1, handle_na="drop"):
return None
-def cramer_v_series(X, target, sample_weight=None, n_jobs=-1, handle_na="drop"):
+def cramer_v_series(X, target, sample_weight=None, n_jobs=1, handle_na="drop"):
"""cramer_v_series computes the weighted Cramer's V statistic for
categorical-categorical association. This is a symmetric coefficient: V(x,y) = V(y,x)
@@ -393,7 +478,7 @@ def cramer_v_series(X, target, sample_weight=None, n_jobs=-1, handle_na="drop"):
sample_weight : array-like of shape (n_samples,), optional
The weight vector, by default None
n_jobs : int, optional
- the number of cores to use for the computation, by default -1
+ the number of cores to use for the computation, by default 1
handle_na : str, optional
either drop rows with na, fill na with 0 or do nothing, by default "drop"
@@ -453,34 +538,32 @@ def _weighted_correlation_ratio(*args):
float
value of the correlation ratio
"""
- # how many levels (predictor)
- n_classes = len(args)
- # convert to float 2-uple d'array
- args = [as_float_array(a) for a in args]
- # compute the total weight per level
+ # Convert to float array and compute weights
+ args = [(np.asarray(a[0], dtype=float), np.asarray(a[1])) for a in args]
weight_per_class = np.array([a[1].sum() for a in args])
- # total weight
- tot_weight = np.sum(weight_per_class)
- # weighted sum of squares
- ss_alldata = sum((a[1] * safe_sqr(a[0])).sum(axis=0) for a in args)
- # list of weighted sums
- sums_args = [np.asarray((a[0] * a[1]).sum(axis=0)) for a in args]
+ tot_weight = weight_per_class.sum()
+
+ # Weighted sum of squares and list of weighted sums
+ ss_alldata = sum((a[1] * np.square(a[0])).sum(axis=0) for a in args)
+ sums_args = [np.sum(a[0] * a[1], axis=0) for a in args]
square_of_sums_alldata = sum(sums_args) ** 2
- square_of_sums_args = [s**2 for s in sums_args]
- sstot = ss_alldata - square_of_sums_alldata / float(tot_weight)
- ssbn = 0.0
- for k, _ in enumerate(args):
- ssbn += square_of_sums_args[k] / weight_per_class[k]
- ssbn -= square_of_sums_alldata / float(tot_weight)
+
+ # Total sum of squares and between-classes sum of squares
+ sstot = ss_alldata - square_of_sums_alldata / tot_weight
+ ssbn = sum(np.square(s) / w for s, w in zip(sums_args, weight_per_class))
+ ssbn -= square_of_sums_alldata / tot_weight
+
+ # Handle constant features
constant_features_idx = np.where(sstot == 0.0)[0]
- if np.nonzero(ssbn)[0].size != ssbn.size and constant_features_idx.size:
+ if np.any(ssbn) and constant_features_idx.size:
warnings.warn("Features %s are constant." % constant_features_idx, UserWarning)
- etasq = ssbn / sstot
- # flatten matrix to vector in sparse case
- etasq = np.asarray(etasq).ravel()
- return np.sqrt(etasq)
+ # Correlation Ratio calculation
+ etasq = np.divide(ssbn, sstot, out=np.zeros_like(ssbn), where=sstot != 0)
+ return np.sqrt(etasq).ravel()
+
+@symmetric_function
def correlation_ratio(x, y, sample_weight=None, as_frame=False):
"""Compute the weighted correlation ratio. The association between a continuous predictor (y)
and a categorical predictor (x). It can be weighted.
@@ -501,43 +584,46 @@ def correlation_ratio(x, y, sample_weight=None, as_frame=False):
float
value of the correlation ratio
"""
+ if not (isinstance(x, pd.Series) and isinstance(y, pd.Series)):
+ raise TypeError(
+ f"Both x and y must be pandas Series. The type of x: {type(x)} and The type of y: {type(y)}"
+ )
+
if sample_weight is None:
sample_weight = np.ones_like(y)
- # one 2-uple per level of the categorical feature x
+ # Determine the categorical and continuous variables
if x.dtype in ["category", "object", "bool"]:
- args = [
- (
- y[safe_mask(y, x == k)],
- sample_weight[safe_mask(sample_weight, x == k)],
- )
- for k in np.unique(x)
- ]
+ categorical, continuous = x, y
elif y.dtype in ["category", "object", "bool"]:
- args = [
- (
- x[safe_mask(x, y == k)],
- sample_weight[safe_mask(sample_weight, y == k)],
- )
- for k in np.unique(y)
- ]
+ categorical, continuous = y, x
else:
raise TypeError(
- "one of the two series should be categorical/object and the other numerical"
+ "One of the series must be categorical and the other numerical."
)
+ # Prepare arguments for the weighted correlation ratio calculation
+ unique_categories = np.unique(categorical)
+ args = [
+ (continuous[categorical == category], sample_weight[categorical == category])
+ for category in unique_categories
+ ]
+
+ # Compute the weighted correlation ratio
+ v = _weighted_correlation_ratio(*args)[0]
+
+ # Format the result
if as_frame:
- x_name = x.name if isinstance(x, pd.Series) else "var"
- y_name = y.name if isinstance(y, pd.Series) else "target"
- v = _weighted_correlation_ratio(*args)[0]
+ x_name = x.name if x.name else "var1"
+ y_name = y.name if y.name else "var2"
return pd.DataFrame(
{"row": [x_name, y_name], "col": [y_name, x_name], "val": [v, v]}
)
else:
- return _weighted_correlation_ratio(*args)[0]
+ return v
-def correlation_ratio_matrix(X, sample_weight=None, n_jobs=-1, handle_na="drop"):
+def correlation_ratio_matrix(X, sample_weight=None, n_jobs=1, handle_na="drop"):
"""correlation_ratio_matrix computes the weighted Correlation Ratio for
categorical-numerical association. This is a symmetric coefficient: CR(x,y) = CR(y,x)
@@ -552,7 +638,7 @@ def correlation_ratio_matrix(X, sample_weight=None, n_jobs=-1, handle_na="drop")
sample_weight : array-like of shape (n_samples,), optional
The weight vector, by default None
n_jobs : int, optional
- the number of cores to use for the computation, by default -1
+ the number of cores to use for the computation, by default 1
handle_na : str, optional
either drop rows with na, fill na with 0 or do nothing, by default "drop"
@@ -588,16 +674,14 @@ def correlation_ratio_matrix(X, sample_weight=None, n_jobs=-1, handle_na="drop")
df=X,
comb_list=comb_list,
sample_weight=sample_weight,
- n_jobs=-1,
+ n_jobs=n_jobs,
)
return lst
else:
return None
-def correlation_ratio_series(
- X, target, sample_weight=None, n_jobs=-1, handle_na="drop"
-):
+def correlation_ratio_series(X, target, sample_weight=None, n_jobs=1, handle_na="drop"):
"""correlation_ratio_series computes the weighted correlation ration for
categorical-numerical association. This is a symmetric coefficient: CR(x,y) = CR(y,x)
@@ -756,6 +840,7 @@ def wspearman(x, y, w):
return wcorr(wrank(x, w), wrank(y, w), w)
+@symmetric_function
def weighted_corr(x, y, sample_weight=None, as_frame=False, method="pearson"):
"""weighted_corr computes the weighted correlation coefficient (Pearson or Spearman)
@@ -777,6 +862,8 @@ def weighted_corr(x, y, sample_weight=None, as_frame=False, method="pearson"):
float or pd.DataFrame
weighted correlation coefficient
"""
+ if sample_weight is None:
+ sample_weight = np.ones_like(y)
if method == "pearson":
c = wcorr(x, y, sample_weight)
@@ -794,7 +881,7 @@ def weighted_corr(x, y, sample_weight=None, as_frame=False, method="pearson"):
def wcorr_series(
- X, target, sample_weight=None, n_jobs=-1, handle_na="drop", method="pearson"
+ X, target, sample_weight=None, n_jobs=1, handle_na="drop", method="pearson"
):
"""wcorr_series computes the weighted correlation coefficient (Pearson or Spearman) for
continuous-continuous association. This is an symmetric coefficient: corr(x,y) = corr(y,x)
@@ -812,7 +899,7 @@ def wcorr_series(
sample_weight : array-like of shape (n_samples,), optional
The weight vector, by default None
n_jobs : int, optional
- the number of cores to use for the computation, by default -1
+ the number of cores to use for the computation, by default 1
handle_na : str, optional
either drop rows with na, fill na with 0 or do nothing, by default "drop"
method : str
@@ -849,7 +936,7 @@ def wcorr_series(
return None
-def wcorr_matrix(X, sample_weight=None, n_jobs=-1, handle_na="drop", method="pearson"):
+def wcorr_matrix(X, sample_weight=None, n_jobs=1, handle_na="drop", method="pearson"):
"""wcorr_matrix computes the weighted correlation statistic for
(Pearson or Spearman) for continuous-continuous association.
This is an symmetric coefficient: corr(x,y) = corr(y,x)
@@ -902,7 +989,7 @@ def wcorr_matrix(X, sample_weight=None, n_jobs=-1, handle_na="drop", method="pea
df=X,
comb_list=comb_list,
sample_weight=sample_weight,
- n_jobs=-1,
+ n_jobs=n_jobs,
)
return lst
else:
@@ -977,180 +1064,112 @@ def association_series(
target,
features=None,
sample_weight=None,
- nom_nom_assoc="theil",
- num_num_assoc="pearson",
- nom_num_assoc="correlation_ratio",
+ nom_nom_assoc=weighted_theils_u,
+ num_num_assoc=partial(weighted_corr, method="spearman"),
+ nom_num_assoc=correlation_ratio,
normalize=False,
- n_jobs=-1,
+ n_jobs=1,
handle_na="drop",
):
- """association_series computes the association matrix for cont-cont, cat-cont and cat-cat.
- predictors. The weighted correlation matrix is used for the cont-cont predictors.
- The correlation ratio is used between cont-cat predictors and either the Cramer's V or Theil's U
- matrix for cat-cat predictors. The Pearson or Spearman correlation coefficient is used for
- the cont-cont association.
+ """
+ Computes the association series for different types of predictors.
+
+ This function calculates the association between the specified `target` and other predictors in `X`.
+ It supports different types of associations: nominal-nominal, numerical-numerical, and nominal-numerical.
Parameters
----------
- X : array-like of shape (n_samples, n_features)
- predictor dataframe
+ X : array-like, shape (n_samples, n_features)
+ Predictor dataframe.
target : str or int
- the predictor name or index with which to compute association
+ The predictor name or index with which to compute the association.
features : list of str, optional
- list of features with which to compute the association
- sample_weight : array-like of shape (n_samples,), optional
- The weight vector, by default None
- nom_nom_assoc : str or callable
- If callable, a function which receives two `pd.Series` (and optionally a weight array) and returns a single number.
- If string, name of nominal-nominal (categorical-categorical) association to use.
- Options are 'cramer' for Cramer's V or `theil` for Theil's U. If 'theil',
- heat-map columns are the provided information (U = U(row|col)).
- num_num_assoc : str or callable
- If callable, a function which receives two `pd.Series` and returns a single number.
- If string, name of numerical-numerical association to use. Options are 'pearson'
- for Pearson's R, 'spearman' for Spearman's R.
- nom_num_assoc : str or callable
- If callable, a function which receives two `pd.Series` and returns a single number.
- If string, name of nominal-numerical association to use. Options are 'correlation_ratio'
- for correlation ratio
- normalize : bool
- either to normalize or not the scores
+ List of features with which to compute the association. If None, all features in X are used.
+ sample_weight : array-like, shape (n_samples,), optional
+ The weight vector, by default None.
+ nom_nom_assoc : callable
+ Function to compute the nominal-nominal (categorical-categorical) association.
+ It should take two pd.Series and an optional weight array, and return a single number.
+ num_num_assoc : callable
+ Function to compute the numerical-numerical association.
+ It should take two pd.Series and return a single number.
+ nom_num_assoc : callable
+ Function to compute the nominal-numerical association.
+ It should take two pd.Series and return a single number.
+ normalize : bool, optional
+ Whether to normalize the scores or not. If True, scores are normalized to the range [0, 1].
n_jobs : int, optional
- the number of cores to use for the computation, by default -1
+ The number of cores to use for the computation. The default, -1, uses all available cores.
handle_na : str, optional
- either drop rows with na, fill na with 0 or do nothing, by default "drop"
+ How to handle NA values. Options are 'drop', 'fill', and None. The default, 'drop', drops rows with NA values.
Returns
-------
pd.Series
- a series with all the association values with the target column
+ A series with all the association values with the target column, sorted in descending order.
Raises
------
TypeError
- if features is not None and is not a list of strings
+ If `features` is provided but is not a list of strings.
+
+ Examples
+ --------
+ >>> import pandas as pd
+ >>> from sklearn import datasets
+ >>> iris = datasets.load_iris()
+ >>> X = pd.DataFrame(iris.data, columns=iris.feature_names)
+ >>> association_series(X, 'sepal length (cm)', num_num_assoc=my_num_num_function)
+
+ Notes
+ -----
+ The function dynamically selects the appropriate association method based on the data types
+ of the target and other predictors. For numerical-numerical associations,
+ it uses `num_num_assoc`; for nominal-nominal, `nom_nom_assoc`; and for nominal-numerical, `nom_num_assoc`.
"""
- # sanity checks
+ # Input validation and preprocessing
X, sample_weight = _check_association_input(X, sample_weight, handle_na)
-
- if features and is_list_of_str(features):
+ if features is not None:
+ if not all(isinstance(f, str) for f in features):
+ raise TypeError("Features must be a list of strings.")
data = X[features + [target]]
- elif features and (not is_list_of_str(features)):
- raise TypeError("features is not a list of strings")
- elif features is None:
- data = X
-
- dtypes_dic = create_dtype_dict(X)
-
- # only numeric, NaN already checked, not repeating the process
- if X.dtypes.map(is_numeric_dtype).all():
- if callable(num_num_assoc):
- return _callable_association_series_fn(
- assoc_fn=num_num_assoc,
- X=data,
- target=target,
- sample_weight=sample_weight,
- n_jobs=n_jobs,
- kind="num-num",
- )
- else:
- return wcorr_series(
- data,
- target,
- sample_weight,
- n_jobs,
- handle_na=None,
- method=num_num_assoc,
- )
+ else:
+ data = X.copy()
- # only categorical (here understood as no numerical columns)
- if not X.dtypes.map(is_numeric_dtype).any():
- if callable(nom_nom_assoc):
- return _callable_association_series_fn(
- assoc_fn=nom_nom_assoc,
- X=data,
- target=target,
- sample_weight=sample_weight,
- n_jobs=n_jobs,
- kind="nom-nom",
- )
- elif nom_nom_assoc == "theil":
- return theils_u_series(data, target, sample_weight, n_jobs, handle_na=None)
- elif nom_nom_assoc == "cramer":
- return cramer_v_series(data, target, sample_weight, n_jobs, handle_na=None)
+ # Determine the data types
+ is_numeric = pd.api.types.is_numeric_dtype
+ numeric_cols = data.select_dtypes(include=[np.number]).columns.tolist()
+ categorical_cols = data.select_dtypes(exclude=[np.number]).columns.tolist()
- # cat-num
- if callable(nom_num_assoc):
+ # Compute associations based on data types
+ if all(is_numeric(data[col]) for col in data.columns):
assoc_series = _callable_association_series_fn(
- assoc_fn=nom_num_assoc,
- X=data,
- target=target,
- sample_weight=sample_weight,
- n_jobs=n_jobs,
- kind="nom-num",
+ num_num_assoc, data, target, sample_weight, n_jobs, "num-num"
+ )
+ elif all(not is_numeric(data[col]) for col in data.columns):
+ assoc_series = _callable_association_series_fn(
+ nom_nom_assoc, data, target, sample_weight, n_jobs, "nom-nom"
)
else:
- assoc_series = correlation_ratio_series(
- data, target, sample_weight, n_jobs, handle_na=None
+ assoc_series = _callable_association_series_fn(
+ nom_num_assoc, data, target, sample_weight, n_jobs, "nom-num"
)
- if normalize:
- assoc_series = (assoc_series - assoc_series.min()) / np.ptp(assoc_series)
-
- # cat-cat
- if dtypes_dic[target] == "cat":
- if callable(nom_nom_assoc):
+ # Additional association for target-specific types
+ if is_numeric(data[target]):
assoc_series_complement = _callable_association_series_fn(
- assoc_fn=nom_nom_assoc,
- X=data,
- target=target,
- sample_weight=sample_weight,
- n_jobs=n_jobs,
- kind="nom-nom",
+ num_num_assoc, data, target, sample_weight, n_jobs, "num-num"
)
- elif nom_nom_assoc == "theil":
- assoc_series_complement = theils_u_series(
- data, target, sample_weight, n_jobs, handle_na=None
- )
- else:
- assoc_series_complement = cramer_v_series(
- data, target, sample_weight, n_jobs, handle_na=None
- )
-
- if normalize:
- assoc_series_complement = (
- assoc_series_complement - assoc_series_complement.min()
- ) / np.ptp(assoc_series_complement)
-
- assoc_series = pd.concat([assoc_series, assoc_series_complement])
-
- # num-num
- if dtypes_dic[target] == "num":
- if callable(num_num_assoc):
+ assoc_series = pd.concat([assoc_series, assoc_series_complement])
+ elif not is_numeric(data[target]):
assoc_series_complement = _callable_association_series_fn(
- assoc_fn=num_num_assoc,
- X=data,
- target=target,
- sample_weight=sample_weight,
- n_jobs=n_jobs,
- kind="num-num",
- )
- else:
- assoc_series_complement = wcorr_series(
- data,
- target,
- sample_weight,
- n_jobs,
- handle_na=None,
- method=num_num_assoc,
+ nom_nom_assoc, data, target, sample_weight, n_jobs, "nom-nom"
)
+ assoc_series = pd.concat([assoc_series, assoc_series_complement])
- if normalize:
- assoc_series_complement = (
- assoc_series_complement - assoc_series_complement.min()
- ) / np.ptp(assoc_series_complement)
-
- assoc_series = pd.concat([assoc_series, assoc_series_complement])
+ # Normalize if required
+ if normalize:
+ assoc_series = (assoc_series - assoc_series.min()) / np.ptp(assoc_series)
return assoc_series.sort_values(ascending=False)
@@ -1158,124 +1177,87 @@ def association_series(
def association_matrix(
X,
sample_weight=None,
- nom_nom_assoc="theil",
- num_num_assoc="pearson",
- nom_num_assoc="correlation_ratio",
- n_jobs=-1,
+ nom_nom_assoc=weighted_theils_u,
+ num_num_assoc=weighted_corr,
+ nom_num_assoc=correlation_ratio,
+ n_jobs=1,
handle_na="drop",
- nom_nom_comb=None,
- num_num_comb=None,
- nom_num_comb=None,
):
- """association_matrix computes the association matrix for cont-cont, cat-cont and cat-cat.
- predictors. The weighted correlation matrix is used for the cont-cont predictors.
- The correlation ratio is used between cont-cat predictors and either the Cramer's V or Theil's U
- matrix for cat-cat predictors.
-
- The association matrix is not symmetric is Theil is used. The obeservations might be weighted.
+ """
+ Computes the association matrix for continuous-continuous, categorical-continuous,
+ and categorical-categorical predictors using specified callable functions.
Parameters
----------
X : array-like of shape (n_samples, n_features)
- predictor dataframe
+ Predictor dataframe.
sample_weight : array-like of shape (n_samples,), optional
- The weight vector, by default None
- nom_nom_assoc : str or callable
- If callable, a function which receives two `pd.Series` (and optionally a weight array) and returns a single number.
- If string, name of nominal-nominal (categorical-categorical) association to use.
- Options are 'cramer' for Cramer's V or `theil` for Theil's U. If 'theil',
- heat-map columns are the provided information (U = U(row|col)).
- num_num_assoc : str or callable
- If callable, a function which receives two `pd.Series` and returns a single number.
- If string, name of numerical-numerical association to use. Options are 'pearson'
- for Pearson's R, 'spearman' for Spearman's R.
- nom_num_assoc : str or callable
- If callable, a function which receives two `pd.Series` and returns a single number.
- If string, name of nominal-numerical association to use. Options are 'correlation_ratio'
- for correlation ratio
+ The weight vector, by default None.
+ nom_nom_assoc : callable
+ Function to compute the categorical-categorical association.
+ num_num_assoc : callable
+ Function to compute the numerical-numerical association.
+ nom_num_assoc : callable
+ Function to compute the categorical-numerical association.
n_jobs : int, optional
- the number of cores to use for the computation, by default -1
+ The number of cores to use for the computation, by default 1.
handle_na : str, optional
- either drop rows with na, fill na with 0 or do nothing, by default "drop"
- nom_nom_comb_list : list of 2-uple of strings
- Pairs of column names corresponding to the entries for nom_nom associations.
- If asymmetrical association, take care of providing an exhaustive list of column name pairs.
- num_num_comb_list : list of 2-uple of strings
- Pairs of column names corresponding to the entries for num_num associations
- nom_num_comb_list : list of 2-uple of strings
- Pairs of column names corresponding to the entries for nom_num associations
+ How to handle NA values ('drop', 'fill', or None), by default "drop".
Returns
-------
pd.DataFrame
- the association matrix
+ The association matrix.
"""
- # sanity checks
+ # Input validation and preprocessing
X, sample_weight = _check_association_input(X, sample_weight, handle_na)
dtypes_dic = create_dtype_dict(X, dic_keys="dtypes")
- # Cramer's V only for categorical columns
- # in GLM supposed to be all the columns
n_cat_cols = len(dtypes_dic["cat"])
n_num_cols = len(dtypes_dic["num"])
-
+
df_to_concat = []
- # num-num, NaNs already checked above, not repeating the process
+ # Numerical-Numerical Associations
if n_num_cols >= 2:
- if callable(num_num_assoc):
- w_num_num = _callable_association_matrix_fn(
- assoc_fn=num_num_assoc,
- cols_comb=num_num_comb,
- kind="num-num",
- X=X,
- sample_weight=sample_weight,
- n_jobs=n_jobs,
- )
- else:
- w_num_num = wcorr_matrix(
- X, sample_weight, n_jobs, handle_na=None, method=num_num_assoc
- )
- df_to_concat.append(w_num_num)
-
- # nom-num
- if (n_num_cols >= 1) and (n_cat_cols >= 1):
- if callable(nom_num_assoc):
- w_nom_num = _callable_association_matrix_fn(
- assoc_fn=nom_num_assoc,
- cols_comb=nom_num_comb,
- kind="nom-num",
- X=X,
- sample_weight=sample_weight,
- n_jobs=n_jobs,
- )
- else:
- w_nom_num = correlation_ratio_matrix(X, sample_weight, n_jobs, handle_na=None)
- df_to_concat.append(w_nom_num)
+ w_num_num = _callable_association_matrix_fn(
+ assoc_fn=num_num_assoc,
+ X=X,
+ sample_weight=sample_weight,
+ n_jobs=n_jobs,
+ kind="num-num",
+ )
+ df_to_concat.append(w_num_num)
- # nom-nom
- if n_cat_cols >= 2:
- if callable(nom_nom_assoc):
- w_nom_nom = _callable_association_matrix_fn(
- assoc_fn=nom_nom_assoc,
- cols_comb=nom_nom_comb,
- kind="nom-nom",
- X=X,
- sample_weight=sample_weight,
- n_jobs=n_jobs,
- )
- elif nom_nom_assoc == "cramer":
- w_nom_nom = cramer_v_matrix(X, sample_weight, n_jobs, handle_na=None)
- else:
- w_nom_nom = theils_u_matrix(X, sample_weight, n_jobs, handle_na=None)
- df_to_concat.append(w_nom_nom)
+ # Categorical-Numerical Associations
+ if n_num_cols >= 1 and n_cat_cols >= 1:
+ w_nom_num = _callable_association_matrix_fn(
+ assoc_fn=nom_num_assoc,
+ X=X,
+ sample_weight=sample_weight,
+ n_jobs=n_jobs,
+ kind="nom-num",
+ )
+ df_to_concat.append(w_nom_num)
+ # Categorical-Categorical Associations
+ if n_cat_cols >= 2:
+ w_nom_nom = _callable_association_matrix_fn(
+ assoc_fn=nom_nom_assoc,
+ X=X,
+ sample_weight=sample_weight,
+ n_jobs=n_jobs,
+ kind="nom-nom",
+ )
+ df_to_concat.append(w_nom_nom)
- return pd.concat(df_to_concat, ignore_index=True)
+ return (
+ pd.concat(df_to_concat, ignore_index=True) if df_to_concat else pd.DataFrame()
+ )
def _callable_association_series_fn(
- assoc_fn, X, target, sample_weight=None, n_jobs=-1, kind="nom-nom"
+ assoc_fn, X, target, sample_weight=None, n_jobs=1, kind="nom-nom"
):
"""_callable_association_series_fn private function, utility for computing association series
for a callable custom association
@@ -1291,7 +1273,7 @@ def _callable_association_series_fn(
sample_weight : array-like of shape (n_samples,), optional
The weight vector, by default None
n_jobs : int, optional
- the number of cores to use for the computation, by default -1
+ the number of cores to use for the computation, by default 1
kind : str
kind of association, either 'num-num' or 'nom-nom' or 'nom-num'
@@ -1305,79 +1287,55 @@ def _callable_association_series_fn(
ValueError
if kind is not 'num-num' or 'nom-nom' or 'nom-num'
"""
+ X, sample_weight = _check_association_input(X, sample_weight, handle_na="drop")
+
+ # Validate 'kind' parameter
+ valid_kinds = ["num-num", "nom-nom", "nom-num"]
+ if kind not in valid_kinds:
+ raise ValueError(f"kind must be one of {valid_kinds}")
+
+ # Create dtype dictionaries
col_dtypes_dic = create_dtype_dict(X)
dtypes_dic = create_dtype_dict(X, dic_keys="dtypes")
- if kind == "nom-nom":
- if col_dtypes_dic[target] != "cat":
- raise TypeError("the target column is not categorical")
- nom_cols = dtypes_dic["cat"]
- if nom_cols:
- # define the number of cores
- n_jobs = (
- min(cpu_count(), len(nom_cols))
- if n_jobs == -1
- else min(cpu_count(), n_jobs)
- )
- # parallelize jobs
- _assoc_fn = partial(_compute_series, func_xyw=assoc_fn)
- return parallel_df(
- func=_assoc_fn,
- df=X[nom_cols],
- series=X[target],
- sample_weight=sample_weight,
- n_jobs=n_jobs,
+ # Determine predictor list based on 'kind'
+ if kind in ["nom-nom", "nom-num"]:
+ if kind == "nom-nom" and col_dtypes_dic[target] != "cat":
+ raise TypeError(
+ "Target column is not categorical for 'nom-nom' association"
)
- else:
- return None
+ pred_list = (
+ dtypes_dic["cat"]
+ if kind == "nom-nom"
+ else dtypes_dic["num"]
+ if col_dtypes_dic[target] == "cat"
+ else dtypes_dic["cat"]
+ )
+ else: # kind == 'num-num'
+ pred_list = dtypes_dic["num"]
- elif kind == "nom-num":
- if col_dtypes_dic[target] == "cat":
- # if the target is categorical, pick only num predictors
- pred_list = dtypes_dic["num"]
- else:
- # if the target is numerical, the 2nd pred should be categorical
- pred_list = dtypes_dic["cat"]
-
- if pred_list:
- # define the number of cores
- n_jobs = (
- min(cpu_count(), len(pred_list))
- if n_jobs == -1
- else min(cpu_count(), n_jobs)
- )
- # parallelize jobs
- _assoc_fn = partial(_compute_series, func_xyw=assoc_fn)
- return parallel_df(
- func=_assoc_fn,
- df=X[pred_list],
- series=X[target],
- sample_weight=sample_weight,
- n_jobs=n_jobs,
- )
- else:
- return None
-
- elif kind == "num-num":
- num_cols = dtypes_dic["num"]
- if num_cols:
- # parallelize jobs
- _assoc_fn = partial(_compute_series, func_xyw=assoc_fn)
- return parallel_df(
- func=_assoc_fn,
- df=X[num_cols],
- series=X[target],
- sample_weight=sample_weight,
- n_jobs=n_jobs,
- )
- else:
- return None
- else:
- raise ValueError("kind can be 'num-num' or 'nom-num' or 'nom-nom'")
+ # Return None if no predictors are available
+ if not pred_list:
+ return None
+
+ # Define the number of cores
+ n_jobs = (
+ min(cpu_count(), len(pred_list)) if n_jobs == -1 else min(cpu_count(), n_jobs)
+ )
+
+ # Setup parallel computation
+ _assoc_fn = partial(_compute_series, func_xyw=assoc_fn)
+ return parallel_df(
+ func=_assoc_fn,
+ df=X[pred_list],
+ series=X[target],
+ sample_weight=sample_weight,
+ n_jobs=n_jobs,
+ )
def _callable_association_matrix_fn(
- assoc_fn, X, sample_weight=None, n_jobs=-1, kind="nom-nom", cols_comb=None
+ assoc_fn, X, sample_weight=None, n_jobs=1, kind="nom-nom", cols_comb=None
):
"""_callable_association_matrix_fn private function, utility for computing association matrix
for a callable custom association
@@ -1402,47 +1360,44 @@ def _callable_association_matrix_fn(
pd.DataFrame
the association matrix
"""
+ # Validate 'kind' parameter
+ valid_kinds = ["num-num", "nom-nom", "nom-num"]
+ if kind not in valid_kinds:
+ raise ValueError(f"kind must be one of {valid_kinds}")
+
+ # Create dtype dictionaries
dtypes_dic = create_dtype_dict(X, dic_keys="dtypes")
+ # Determine column combinations based on 'kind' and 'cols_comb'
if cols_comb is None:
if kind == "num-num":
selected_cols = dtypes_dic["num"]
+ cols_comb = create_col_combinations(assoc_fn, selected_cols)
elif kind == "nom-nom":
selected_cols = dtypes_dic["cat"]
+ cols_comb = create_col_combinations(assoc_fn, selected_cols)
elif kind == "nom-num":
- cat_cols = dtypes_dic["cat"]
- num_cols = dtypes_dic["num"]
- if cat_cols and num_cols:
- # explicitely store the unique 2-combinations of column names
- # the first one should be the categorical predictor
- selected_cols = list(product(cat_cols, num_cols))
- else:
- selected_cols = None
-
- if selected_cols:
- # explicitely store the unique 2-combinations of column names
- cols_comb = [comb for comb in combinations(selected_cols, 2)]
- _assoc_fn = partial(_compute_matrix_entries, func_xyw=assoc_fn)
- assoc = parallel_matrix_entries(
- func=_assoc_fn,
- df=X,
- comb_list=cols_comb,
- sample_weight=sample_weight,
- n_jobs=n_jobs,
- )
+ # cols_comb = create_col_combinations(assoc_fn, selected_cols)
+ cols_comb = list(product(dtypes_dic["cat"], dtypes_dic["num"]))
- else:
- assoc = None
- else:
- _assoc_fn = partial(_compute_matrix_entries, func_xyw=assoc_fn)
- assoc = parallel_matrix_entries(
- func=_assoc_fn,
- df=X,
- comb_list=cols_comb,
- sample_weight=sample_weight,
- n_jobs=n_jobs,
- )
- return assoc
+ # Return None if no column combinations are available
+ if not cols_comb:
+ return None
+
+ # Define the number of cores
+ n_jobs = (
+ min(cpu_count(), len(cols_comb)) if n_jobs == -1 else min(cpu_count(), n_jobs)
+ )
+
+ # Setup parallel computation
+ _assoc_fn = partial(_compute_matrix_entries, func_xyw=assoc_fn)
+ return parallel_matrix_entries(
+ func=_assoc_fn,
+ df=X,
+ comb_list=cols_comb,
+ sample_weight=sample_weight,
+ n_jobs=n_jobs,
+ )
################################
@@ -1551,7 +1506,7 @@ def f_cat_regression(x, y, sample_weight=None, as_frame=False):
return f_oneway_weighted(*args)[0]
-def f_cat_regression_parallel(X, y, sample_weight=None, n_jobs=-1, handle_na="drop"):
+def f_cat_regression_parallel(X, y, sample_weight=None, n_jobs=1, handle_na="drop"):
"""f_cat_regression_parallel computes the weighted ANOVA F-value for the provided categorical predictors
using parallelization of the code (continuous target, categorical predictor).
@@ -1564,7 +1519,7 @@ def f_cat_regression_parallel(X, y, sample_weight=None, n_jobs=-1, handle_na="dr
sample_weight : array-like of shape (n_samples,), optional
The weight vector, by default None
n_jobs : int, optional
- the number of cores to use for the computation, by default -1
+ the number of cores to use for the computation, by default 1
handle_na : str, optional
either drop rows with na, fill na with 0 or do nothing, by default "drop"
diff --git a/src/arfs/feature_selection/mrmr.py b/src/arfs/feature_selection/mrmr.py
index 11b74d1..de4b087 100644
--- a/src/arfs/feature_selection/mrmr.py
+++ b/src/arfs/feature_selection/mrmr.py
@@ -110,7 +110,7 @@ def __init__(
denominator_func=np.mean,
only_same_domain=False,
return_scores=False,
- n_jobs=-1,
+ n_jobs=1,
show_progress=True,
):
self.n_features_to_select = n_features_to_select
diff --git a/src/arfs/feature_selection/unsupervised.py b/src/arfs/feature_selection/unsupervised.py
index 318842d..a5b85e4 100644
--- a/src/arfs/feature_selection/unsupervised.py
+++ b/src/arfs/feature_selection/unsupervised.py
@@ -27,7 +27,14 @@
# ARFS
from .base import BaseThresholdSelector
from ..utils import create_dtype_dict
-from ..association import association_matrix, xy_to_matrix, plot_association_matrix
+from ..association import (
+ association_matrix,
+ xy_to_matrix,
+ plot_association_matrix,
+ weighted_theils_u,
+ weighted_corr,
+ correlation_ratio,
+)
from ..preprocessing import OrdinalEncoderPandas
@@ -296,10 +303,10 @@ def __init__(
self,
threshold=0.80,
method="association",
- n_jobs=-1,
- nom_nom_assoc="theil",
- num_num_assoc="spearman",
- nom_num_assoc="correlation_ratio",
+ n_jobs=1,
+ nom_nom_assoc=weighted_theils_u,
+ num_num_assoc=weighted_corr,
+ nom_num_assoc=correlation_ratio,
):
self.threshold = threshold
self.method = method
@@ -444,7 +451,7 @@ def _recursive_collinear_elimination(association_matrix, threshold):
while True:
most_collinear_feature, to_drop = _most_collinear(dum, threshold)
-
+
# Break if no more features to drop
if not to_drop:
break
@@ -453,4 +460,4 @@ def _recursive_collinear_elimination(association_matrix, threshold):
most_collinear_features.append(most_collinear_feature)
dum = dum.drop(columns=most_collinear_feature, index=most_collinear_feature)
- return most_collinear_features
\ No newline at end of file
+ return most_collinear_features
diff --git a/src/arfs/parallel.py b/src/arfs/parallel.py
index c5002c4..cc61b3e 100644
--- a/src/arfs/parallel.py
+++ b/src/arfs/parallel.py
@@ -39,20 +39,23 @@ def parallel_matrix_entries(func, df, comb_list, sample_weight=None, n_jobs=-1):
pd.DataFrame
concatenated results into a single pandas DF
"""
+ # Determining the number of jobs
+ n_jobs = cpu_count() if n_jobs == -1 else min(cpu_count(), n_jobs)
if n_jobs == 1:
- return func(X=df, sample_weight=sample_weight, comb_list=comb_list)
+ lst = func(X=df, sample_weight=sample_weight, comb_list=comb_list)
+ return pd.concat(lst, ignore_index=True).sort_values("val", ascending=False)
- n_jobs = (
- min(cpu_count(), len(df.columns)) if n_jobs == -1 else min(cpu_count(), n_jobs)
- )
comb_chunks = np.array_split(comb_list, n_jobs)
lst = Parallel(n_jobs=n_jobs)(
delayed(func)(X=df, sample_weight=sample_weight, comb_list=comb_chunk)
for comb_chunk in comb_chunks
)
- # return flatten list of pandas DF
- return pd.concat(list(chain(*lst)), ignore_index=True)
+ # Directly return the single DataFrame if lst contains only one element
+ if len(lst) == 1:
+ return lst[0]
+ else:
+ return pd.concat(list(chain(*lst)), ignore_index=True)
def parallel_df(func, df, series, sample_weight=None, n_jobs=-1):
@@ -77,19 +80,25 @@ def parallel_df(func, df, series, sample_weight=None, n_jobs=-1):
pd.DataFrame
concatenated results into a single pandas DF
"""
+ # Determining the number of jobs
+ n_jobs = cpu_count() if n_jobs == -1 else min(cpu_count(), n_jobs)
if n_jobs == 1:
- return func(df, series, sample_weight).sort_values(ascending=False)
+ lst = func(df, series, sample_weight).sort_values(ascending=False)
- n_jobs = (
- min(cpu_count(), len(df.columns)) if n_jobs == -1 else min(cpu_count(), n_jobs)
- )
- col_chunks = np.array_split(range(len(df.columns)), n_jobs)
- lst = Parallel(n_jobs=n_jobs)(
- delayed(func)(df.iloc[:, col_chunk], series, sample_weight)
- for col_chunk in col_chunks
- )
- return pd.concat(lst).sort_values(ascending=False)
+ return (
+ pd.concat(lst, ignore_index=True).sort_values("val", ascending=False)
+ if isinstance(lst, list)
+ else lst
+ )
+ else:
+ col_chunks = np.array_split(range(len(df.columns)), n_jobs)
+ lst = Parallel(n_jobs=n_jobs)(
+ delayed(func)(df.iloc[:, col_chunk], series, sample_weight)
+ for col_chunk in col_chunks
+ )
+
+ return pd.concat(lst).sort_values(ascending=False)
def _compute_series(
@@ -157,15 +166,9 @@ def _compute_matrix_entries(
pd.DataFrame
concatenated results into a single pandas DF
"""
- v_df_list = []
- for comb in comb_list:
- v_df_list.append(
- func_xyw(
- x=X[comb[0]],
- y=X[comb[1]],
- sample_weight=sample_weight,
- as_frame=True,
- )
- )
+ v_df_list = [
+ func_xyw(x=X[comb[0]], y=X[comb[1]], sample_weight=sample_weight, as_frame=True)
+ for comb in comb_list
+ ]
return v_df_list
diff --git a/src/arfs/utils.py b/src/arfs/utils.py
index ac77b99..28ed71b 100644
--- a/src/arfs/utils.py
+++ b/src/arfs/utils.py
@@ -37,6 +37,7 @@
# #
#####################
+
def concat_or_group(col, x, max_length=25):
"""
Concatenate unique values from a column or return a group value.
@@ -79,8 +80,12 @@ def concat_or_group(col, x, max_length=25):
>>> {'Category': {1: 'gr_1', 2: 'gr_2', 3: 'E'}}
"""
unique_values = x[col].unique()
- concat_str = " / ".join(map(str, unique_values))
- return concat_str if len(concat_str) < max_length else concat_str[:7] + "/.../" + concat_str[-7:]
+ concat_str = " / ".join(map(str, unique_values))
+ return (
+ concat_str
+ if len(concat_str) < max_length
+ else concat_str[:7] + "/.../" + concat_str[-7:]
+ )
def reset_plot():
@@ -498,7 +503,9 @@ def _get_titanic_data():
# Fetch Titanic data and add random cat and numbers
# Example taken from https://scikit-learn.org/stable/auto_examples/inspection/
# plot_permutation_importance.html#sphx-glr-auto-examples-inspection-plot-permutation-importance-py
- X, y = fetch_openml("titanic", version=1, as_frame=True, return_X_y=True)
+ X, y = fetch_openml(
+ "titanic", version=1, as_frame=True, return_X_y=True, parser="auto"
+ )
rng = np.random.RandomState(seed=42)
nice_guys = ["Rick", "Bender", "Cartman", "Morty", "Fry", "Vador", "Thanos"]
X["random_cat"] = np.random.choice(nice_guys, X.shape[0])