Skip to content

Instantly share code, notes, and snippets.

@brusangues

brusangues/pipeline_catboost.ipynb

Created December 15, 2025 01:48
Show Gist options

  • Select an option

    Learn more about clone URLs
  • Save brusangues/3533b3bc3883bfeeca2c2012e5a47b80 to your computer and use it in GitHub Desktop.

Select an option

Learn more about clone URLs
Save brusangues/3533b3bc3883bfeeca2c2012e5a47b80 to your computer and use it in GitHub Desktop.
Download ZIP
Display the source blob
Display the rendered blob
Raw
pipeline_catboost.ipynb
{
"cells": [
{
"cell_type": "code",
"execution_count": 50,
"id": "36ebe4c8",
"metadata": {},
"outputs": [],
"source": [
"from sklearn.datasets import load_breast_cancer\n",
"\n",
"data = load_breast_cancer(as_frame=True)\n",
"data.keys()\n",
"df = data.frame"
]
},
{
"cell_type": "code",
"execution_count": 51,
"id": "0e5ec84c",
"metadata": {},
"outputs": [],
"source": [
"from sklearn.datasets import fetch_covtype\n",
"\n",
"data = fetch_covtype(as_frame=True)\n",
"data.keys()\n",
"df = data.frame"
]
},
{
"cell_type": "code",
"execution_count": 61,
"id": "b5c86837",
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"Index(['Elevation', 'Aspect', 'Slope', 'Horizontal_Distance_To_Hydrology',\n",
" 'Vertical_Distance_To_Hydrology', 'Horizontal_Distance_To_Roadways',\n",
" 'Hillshade_9am', 'Hillshade_Noon', 'Hillshade_3pm',\n",
" 'Horizontal_Distance_To_Fire_Points', 'Wilderness_Area_0',\n",
" 'Wilderness_Area_1', 'Wilderness_Area_2', 'Wilderness_Area_3',\n",
" 'Soil_Type_0', 'Soil_Type_1', 'Soil_Type_2', 'Soil_Type_3',\n",
" 'Soil_Type_4', 'Soil_Type_5', 'Soil_Type_6', 'Soil_Type_7',\n",
" 'Soil_Type_8', 'Soil_Type_9', 'Soil_Type_10', 'Soil_Type_11',\n",
" 'Soil_Type_12', 'Soil_Type_13', 'Soil_Type_14', 'Soil_Type_15',\n",
" 'Soil_Type_16', 'Soil_Type_17', 'Soil_Type_18', 'Soil_Type_19',\n",
" 'Soil_Type_20', 'Soil_Type_21', 'Soil_Type_22', 'Soil_Type_23',\n",
" 'Soil_Type_24', 'Soil_Type_25', 'Soil_Type_26', 'Soil_Type_27',\n",
" 'Soil_Type_28', 'Soil_Type_29', 'Soil_Type_30', 'Soil_Type_31',\n",
" 'Soil_Type_32', 'Soil_Type_33', 'Soil_Type_34', 'Soil_Type_35',\n",
" 'Soil_Type_36', 'Soil_Type_37', 'Soil_Type_38', 'Soil_Type_39',\n",
" 'Cover_Type'],\n",
" dtype='object')"
]
},
"execution_count": 61,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"df.columns"
]
},
{
"cell_type": "code",
"execution_count": 62,
"id": "cce83c33",
"metadata": {},
"outputs": [],
"source": [
"df.rename(columns={\"Cover_Type\": \"target\"}, inplace=True)"
]
},
{
"cell_type": "code",
"execution_count": 52,
"id": "37120478",
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"(581012, 55)"
]
},
"execution_count": 52,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"df.shape"
]
},
{
"cell_type": "code",
"execution_count": 63,
"id": "d3b943be",
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"target\n",
"2 283301\n",
"1 211840\n",
"3 35754\n",
"7 20510\n",
"6 17367\n",
"5 9493\n",
"4 2747\n",
"Name: count, dtype: int64"
]
},
"execution_count": 63,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"df.target.value_counts()"
]
},
{
"cell_type": "code",
"execution_count": 202,
"id": "8e749d46",
"metadata": {},
"outputs": [],
"source": [
"# Opcional - transformando target em classificação binária\n",
"df.target = df.target.apply(lambda x: int(x >= 4))"
]
},
{
"cell_type": "code",
"execution_count": 209,
"id": "7a486a9e",
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"target\n",
"0 530895\n",
"1 50117\n",
"Name: count, dtype: int64"
]
},
"execution_count": 209,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"df.target.value_counts()"
]
},
{
"cell_type": "code",
"execution_count": 203,
"id": "d048c16f",
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"Index(['elevation', 'aspect', 'slope', 'horizontal_distance_to_hydrology',\n",
" 'vertical_distance_to_hydrology', 'horizontal_distance_to_roadways',\n",
" 'hillshade_9am', 'hillshade_noon', 'hillshade_3pm',\n",
" 'horizontal_distance_to_fire_points', 'wilderness_area_0',\n",
" 'wilderness_area_1', 'wilderness_area_2', 'wilderness_area_3',\n",
" 'soil_type_0', 'soil_type_1', 'soil_type_2', 'soil_type_3',\n",
" 'soil_type_4', 'soil_type_5', 'soil_type_6', 'soil_type_7',\n",
" 'soil_type_8', 'soil_type_9', 'soil_type_10', 'soil_type_11',\n",
" 'soil_type_12', 'soil_type_13', 'soil_type_14', 'soil_type_15',\n",
" 'soil_type_16', 'soil_type_17', 'soil_type_18', 'soil_type_19',\n",
" 'soil_type_20', 'soil_type_21', 'soil_type_22', 'soil_type_23',\n",
" 'soil_type_24', 'soil_type_25', 'soil_type_26', 'soil_type_27',\n",
" 'soil_type_28', 'soil_type_29', 'soil_type_30', 'soil_type_31',\n",
" 'soil_type_32', 'soil_type_33', 'soil_type_34', 'soil_type_35',\n",
" 'soil_type_36', 'soil_type_37', 'soil_type_38', 'soil_type_39',\n",
" 'target'],\n",
" dtype='object')"
]
},
"execution_count": 203,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"df.columns = df.columns.str.lower().str.replace(\" \",\"_\")\n",
"df.columns"
]
},
{
"cell_type": "code",
"execution_count": 204,
"id": "1952689b",
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"df_train.shape=(4183, 55)\n",
"df_eval.shape=(465, 55)\n",
"df_test.shape=(1162, 55)\n"
]
}
],
"source": [
"# Separação dos conjuntos\n",
"\n",
"from sklearn.model_selection import train_test_split as tts\n",
"\n",
"df_sample = df.sample(frac=0.01, random_state=42)\n",
"df_train_eval, df_test = tts(df_sample, test_size=0.2, random_state=42, stratify=df_sample.target)\n",
"df_train, df_eval = tts(df_train_eval, test_size=0.1, random_state=42, stratify=df_train_eval.target)\n",
"\n",
"print(f\"{df_train.shape=}\\n{df_eval.shape=}\\n{df_test.shape=}\")"
]
},
{
"cell_type": "code",
"execution_count": 212,
"id": "7db81aeb",
"metadata": {},
"outputs": [],
"source": [
"# Definindo parâmetros do catboost para usar no resto do pipeline\n",
"params = {\n",
" \"iterations\": 2000,\n",
" \"loss_function\": \"Logloss\", # \"MultiClass\",\n",
" \"eval_metric\": \"F1\", # \"TotalF1\",\n",
" \"random_seed\": 100,\n",
" \"verbose\": 100,\n",
" \"early_stopping_rounds\": 100,\n",
"}"
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "e515191b",
"metadata": {},
"outputs": [
{
"data": {
"application/vnd.jupyter.widget-view+json": {
"model_id": "083cecc8fce2400d89bdb652f6ae43ca",
"version_major": 2,
"version_minor": 0
},
"text/plain": [
"MetricVisualizer(layout=Layout(align_self='stretch', height='500px'))"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"Learning rate set to 0.033379\n",
"0:\tlearn: 0.0976864\ttest: 0.0000000\tbest: 0.0000000 (0)\ttotal: 8.43ms\tremaining: 16.9s\n",
"100:\tlearn: 0.4000000\ttest: 0.3773585\tbest: 0.4150943 (91)\ttotal: 941ms\tremaining: 17.7s\n",
"200:\tlearn: 0.6171429\ttest: 0.4406780\tbest: 0.4482759 (153)\ttotal: 1.75s\tremaining: 15.6s\n",
"300:\tlearn: 0.7033748\ttest: 0.5312500\tbest: 0.5396825 (264)\ttotal: 2.56s\tremaining: 14.5s\n",
"400:\tlearn: 0.7805695\ttest: 0.5454545\tbest: 0.5757576 (318)\ttotal: 3.37s\tremaining: 13.4s\n",
"Stopped by overfitting detector (100 iterations wait)\n",
"\n",
"bestTest = 0.5757575758\n",
"bestIteration = 318\n",
"\n",
"Shrink model to first 319 iterations.\n"
]
},
{
"data": {
"text/plain": [
"<catboost.core.CatBoostClassifier at 0x1a4daf69290>"
]
},
"execution_count": 213,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"# Treinando modelo inicial com todas as features\n",
"from catboost import CatBoostClassifier, Pool, EShapCalcType, EFeaturesSelectionAlgorithm\n",
"\n",
"pool_train = Pool(data=df_train.drop(columns=[\"target\"]), label=df_train[\"target\"])\n",
"pool_eval = Pool(data=df_eval.drop(columns=[\"target\"]), label=df_eval[\"target\"])\n",
"pool_test = Pool(data=df_test.drop(columns=[\"target\"]), label=df_test[\"target\"])\n",
"\n",
"model = CatBoostClassifier(**params)\n",
"\n",
"model.fit(\n",
" pool_train,\n",
" eval_set=pool_eval,\n",
" plot=True\n",
")"
]
},
{
"cell_type": "code",
"execution_count": 214,
"id": "d760142e",
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Selecting 1 features out of 54 using 55 steps\n"
]
},
{
"data": {
"application/vnd.jupyter.widget-view+json": {
"model_id": "4a77eea297c3474a8489436b08b56f48",
"version_major": 2,
"version_minor": 0
},
"text/plain": [
"MetricVisualizer(layout=Layout(align_self='stretch', height='500px'))"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"application/vnd.plotly.v1+json": {
"config": {
"plotlyServerURL": "https://plot.ly"
},
"data": [
{
"line": {
"color": "rgb(51,160,44)"
},
"mode": "lines+markers",
"name": "",
"text": [
"",
"vertical_distance_to_hydrology",
"soil_type_39",
"soil_type_13",
"soil_type_0",
"soil_type_31",
"slope",
"soil_type_2",
"soil_type_15",
"aspect",
"soil_type_12",
"wilderness_area_1",
"soil_type_36",
"soil_type_27",
"soil_type_5",
"soil_type_33",
"soil_type_34",
"soil_type_17",
"hillshade_noon",
"soil_type_26",
"soil_type_6",
"soil_type_10",
"soil_type_19",
"soil_type_8",
"soil_type_7",
"soil_type_14",
"soil_type_20",
"soil_type_24",
"soil_type_18",
"soil_type_16",
"soil_type_25",
"soil_type_35",
"soil_type_22",
"soil_type_28",
"soil_type_23",
"soil_type_21",
"soil_type_1",
"soil_type_11",
"soil_type_4",
"soil_type_32",
"soil_type_30",
"hillshade_9am",
"wilderness_area_2",
"soil_type_3",
"soil_type_29",
"soil_type_37",
"horizontal_distance_to_hydrology",
"soil_type_9",
"horizontal_distance_to_fire_points",
"wilderness_area_3",
"wilderness_area_0",
"soil_type_38",
"horizontal_distance_to_roadways",
"hillshade_3pm"
],
"type": "scatter",
"x": [
0,
1,
2,
3,
4,
5,
6,
7,
8,
9,
10,
11,
12,
13,
14,
15,
16,
17,
18,
19,
20,
21,
22,
23,
24,
25,
26,
27,
28,
29,
30,
31,
32,
33,
34,
35,
36,
37,
38,
39,
40,
41,
42,
43,
44,
45,
46,
47,
48,
49,
50,
51,
52,
53
],
"y": [
0.1477282443515885,
0.14360728665267491,
0.1431814290361284,
0.14298460583655714,
0.14282510476834823,
0.138720506542803,
0.1379514449745828,
0.1373899846007138,
0.13718637249828747,
0.1396625950309679,
0.14072820600529518,
0.14097406332773885,
0.13595778276128634,
0.1346676493523341,
0.13361166554323825,
0.13187604508174725,
0.13120878267479325,
0.13104644007439964,
0.1517044460271607,
0.15171273628228008,
0.1517137533959223,
0.13722506984594268,
0.13210261129011047,
0.1443178728641258,
0.14431787286941414,
0.1458053618311109,
0.14729285079280766,
0.14878033975450441,
0.14132961172824576,
0.14132961172824576,
0.14027958275842997,
0.14027958275843014,
0.14037541750914248,
0.13926717055366758,
0.13797897441609677,
0.14212675332306965,
0.14041561086454277,
0.13757352885677757,
0.13510393460777329,
0.14409481128293467,
0.14084513882448946,
0.12712209783860534,
0.1378335582278886,
0.13478887732444192,
0.13403321273113225,
0.14991679743231506,
0.1350558793736135,
0.14042553951270548,
0.15684548881410224,
0.15089630427895054,
0.15390386308252524,
0.15754301616669614,
0.1922590970216916,
0.194306053544715
]
},
{
"marker": {
"size": 10,
"symbol": "square"
},
"mode": "markers",
"name": "",
"text": [
"",
"soil_type_0",
"soil_type_15",
"wilderness_area_1",
"soil_type_5",
"soil_type_17",
"soil_type_6",
"soil_type_19",
"soil_type_7",
"soil_type_24",
"soil_type_16",
"soil_type_25",
"soil_type_22",
"soil_type_23",
"soil_type_21",
"soil_type_11",
"soil_type_4",
"soil_type_32",
"soil_type_30",
"hillshade_9am",
"wilderness_area_2",
"soil_type_3",
"soil_type_29",
"soil_type_29",
"soil_type_37",
"horizontal_distance_to_hydrology",
"horizontal_distance_to_hydrology",
"soil_type_9",
"soil_type_9",
"horizontal_distance_to_fire_points",
"horizontal_distance_to_fire_points",
"wilderness_area_3",
"wilderness_area_3",
"wilderness_area_3",
"wilderness_area_0",
"wilderness_area_0",
"wilderness_area_0",
"soil_type_38",
"soil_type_38",
"soil_type_38",
"soil_type_38",
"horizontal_distance_to_roadways",
"horizontal_distance_to_roadways",
"horizontal_distance_to_roadways",
"horizontal_distance_to_roadways",
"horizontal_distance_to_roadways",
"horizontal_distance_to_roadways",
"horizontal_distance_to_roadways",
"hillshade_3pm",
"hillshade_3pm",
"hillshade_3pm",
"hillshade_3pm",
"hillshade_3pm",
"hillshade_3pm"
],
"type": "scatter",
"x": [
0,
4,
8,
11,
14,
17,
20,
22,
24,
27,
29,
30,
32,
34,
35,
37,
38,
39,
40,
41,
42,
43,
44,
44,
45,
46,
46,
47,
47,
48,
48,
49,
49,
49,
50,
50,
50,
51,
51,
51,
51,
52,
52,
52,
52,
52,
52,
52,
53,
53,
53,
53,
53,
53
],
"y": [
0.1477282443515885,
0.14282510476834823,
0.13718637249828747,
0.14097406332773885,
0.13361166554323825,
0.13104644007439964,
0.1517137533959223,
0.13210261129011047,
0.14431787286941414,
0.14878033975450441,
0.14132961172824576,
0.14027958275842997,
0.14037541750914248,
0.13797897441609677,
0.14212675332306965,
0.13757352885677757,
0.13510393460777329,
0.14409481128293467,
0.14084513882448946,
0.12712209783860534,
0.1378335582278886,
0.13478887732444192,
0.13403321273113225,
0.13403321273113225,
0.14991679743231506,
0.1350558793736135,
0.1350558793736135,
0.14042553951270548,
0.14042553951270548,
0.15684548881410224,
0.15684548881410224,
0.15089630427895054,
0.15089630427895054,
0.15089630427895054,
0.15390386308252524,
0.15390386308252524,
0.15390386308252524,
0.15754301616669614,
0.15754301616669614,
0.15754301616669614,
0.15754301616669614,
0.1922590970216916,
0.1922590970216916,
0.1922590970216916,
0.1922590970216916,
0.1922590970216916,
0.1922590970216916,
0.1922590970216916,
0.194306053544715,
0.194306053544715,
0.194306053544715,
0.194306053544715,
0.194306053544715,
0.194306053544715
]
},
{
"mode": "text",
"name": "",
"text": [
"",
"4",
"53",
"27",
"14",
"45",
"2",
"16",
"29",
"1",
"26",
"11",
"50",
"41",
"19",
"47",
"48",
"31",
"7",
"40",
"20",
"24",
"33",
"22",
"21",
"28",
"34",
"38",
"32",
"30",
"39",
"49",
"36",
"42",
"37",
"35",
"15",
"25",
"18",
"46",
"44",
"6",
"12",
"17",
"43",
"51",
"3",
"23",
"9",
"13",
"10",
"52",
"5",
"8"
],
"textfont": {
"color": "rgb(51,160,44)",
"family": "sans serif",
"size": 18
},
"textposition": "bottom center",
"type": "scatter",
"visible": false,
"x": [
0,
1,
2,
3,
4,
5,
6,
7,
8,
9,
10,
11,
12,
13,
14,
15,
16,
17,
18,
19,
20,
21,
22,
23,
24,
25,
26,
27,
28,
29,
30,
31,
32,
33,
34,
35,
36,
37,
38,
39,
40,
41,
42,
43,
44,
45,
46,
47,
48,
49,
50,
51,
52,
53
],
"y": [
0.1477282443515885,
0.14360728665267491,
0.1431814290361284,
0.14298460583655714,
0.14282510476834823,
0.138720506542803,
0.1379514449745828,
0.1373899846007138,
0.13718637249828747,
0.1396625950309679,
0.14072820600529518,
0.14097406332773885,
0.13595778276128634,
0.1346676493523341,
0.13361166554323825,
0.13187604508174725,
0.13120878267479325,
0.13104644007439964,
0.1517044460271607,
0.15171273628228008,
0.1517137533959223,
0.13722506984594268,
0.13210261129011047,
0.1443178728641258,
0.14431787286941414,
0.1458053618311109,
0.14729285079280766,
0.14878033975450441,
0.14132961172824576,
0.14132961172824576,
0.14027958275842997,
0.14027958275843014,
0.14037541750914248,
0.13926717055366758,
0.13797897441609677,
0.14212675332306965,
0.14041561086454277,
0.13757352885677757,
0.13510393460777329,
0.14409481128293467,
0.14084513882448946,
0.12712209783860534,
0.1378335582278886,
0.13478887732444192,
0.13403321273113225,
0.14991679743231506,
0.1350558793736135,
0.14042553951270548,
0.15684548881410224,
0.15089630427895054,
0.15390386308252524,
0.15754301616669614,
0.1922590970216916,
0.194306053544715
]
},
{
"mode": "text",
"name": "",
"text": [
"",
"vertical_distance_to_hydrology",
"soil_type_39",
"soil_type_13",
"soil_type_0",
"soil_type_31",
"slope",
"soil_type_2",
"soil_type_15",
"aspect",
"soil_type_12",
"wilderness_area_1",
"soil_type_36",
"soil_type_27",
"soil_type_5",
"soil_type_33",
"soil_type_34",
"soil_type_17",
"hillshade_noon",
"soil_type_26",
"soil_type_6",
"soil_type_10",
"soil_type_19",
"soil_type_8",
"soil_type_7",
"soil_type_14",
"soil_type_20",
"soil_type_24",
"soil_type_18",
"soil_type_16",
"soil_type_25",
"soil_type_35",
"soil_type_22",
"soil_type_28",
"soil_type_23",
"soil_type_21",
"soil_type_1",
"soil_type_11",
"soil_type_4",
"soil_type_32",
"soil_type_30",
"hillshade_9am",
"wilderness_area_2",
"soil_type_3",
"soil_type_29",
"soil_type_37",
"horizontal_distance_to_hydrology",
"soil_type_9",
"horizontal_distance_to_fire_points",
"wilderness_area_3",
"wilderness_area_0",
"soil_type_38",
"horizontal_distance_to_roadways",
"hillshade_3pm"
],
"textfont": {
"color": "rgb(51,160,44)",
"family": "sans serif",
"size": 18
},
"textposition": "bottom center",
"type": "scatter",
"visible": false,
"x": [
0,
1,
2,
3,
4,
5,
6,
7,
8,
9,
10,
11,
12,
13,
14,
15,
16,
17,
18,
19,
20,
21,
22,
23,
24,
25,
26,
27,
28,
29,
30,
31,
32,
33,
34,
35,
36,
37,
38,
39,
40,
41,
42,
43,
44,
45,
46,
47,
48,
49,
50,
51,
52,
53
],
"y": [
0.1477282443515885,
0.14360728665267491,
0.1431814290361284,
0.14298460583655714,
0.14282510476834823,
0.138720506542803,
0.1379514449745828,
0.1373899846007138,
0.13718637249828747,
0.1396625950309679,
0.14072820600529518,
0.14097406332773885,
0.13595778276128634,
0.1346676493523341,
0.13361166554323825,
0.13187604508174725,
0.13120878267479325,
0.13104644007439964,
0.1517044460271607,
0.15171273628228008,
0.1517137533959223,
0.13722506984594268,
0.13210261129011047,
0.1443178728641258,
0.14431787286941414,
0.1458053618311109,
0.14729285079280766,
0.14878033975450441,
0.14132961172824576,
0.14132961172824576,
0.14027958275842997,
0.14027958275843014,
0.14037541750914248,
0.13926717055366758,
0.13797897441609677,
0.14212675332306965,
0.14041561086454277,
0.13757352885677757,
0.13510393460777329,
0.14409481128293467,
0.14084513882448946,
0.12712209783860534,
0.1378335582278886,
0.13478887732444192,
0.13403321273113225,
0.14991679743231506,
0.1350558793736135,
0.14042553951270548,
0.15684548881410224,
0.15089630427895054,
0.15390386308252524,
0.15754301616669614,
0.1922590970216916,
0.194306053544715
]
}
],
"layout": {
"showlegend": false,
"template": {
"data": {
"bar": [
{
"error_x": {
"color": "#2a3f5f"
},
"error_y": {
"color": "#2a3f5f"
},
"marker": {
"line": {
"color": "#E5ECF6",
"width": 0.5
},
"pattern": {
"fillmode": "overlay",
"size": 10,
"solidity": 0.2
}
},
"type": "bar"
}
],
"barpolar": [
{
"marker": {
"line": {
"color": "#E5ECF6",
"width": 0.5
},
"pattern": {
"fillmode": "overlay",
"size": 10,
"solidity": 0.2
}
},
"type": "barpolar"
}
],
"carpet": [
{
"aaxis": {
"endlinecolor": "#2a3f5f",
"gridcolor": "white",
"linecolor": "white",
"minorgridcolor": "white",
"startlinecolor": "#2a3f5f"
},
"baxis": {
"endlinecolor": "#2a3f5f",
"gridcolor": "white",
"linecolor": "white",
"minorgridcolor": "white",
"startlinecolor": "#2a3f5f"
},
"type": "carpet"
}
],
"choropleth": [
{
"colorbar": {
"outlinewidth": 0,
"ticks": ""
},
"type": "choropleth"
}
],
"contour": [
{
"colorbar": {
"outlinewidth": 0,
"ticks": ""
},
"colorscale": [
[
0,
"#0d0887"
],
[
0.1111111111111111,
"#46039f"
],
[
0.2222222222222222,
"#7201a8"
],
[
0.3333333333333333,
"#9c179e"
],
[
0.4444444444444444,
"#bd3786"
],
[
0.5555555555555556,
"#d8576b"
],
[
0.6666666666666666,
"#ed7953"
],
[
0.7777777777777778,
"#fb9f3a"
],
[
0.8888888888888888,
"#fdca26"
],
[
1,
"#f0f921"
]
],
"type": "contour"
}
],
"contourcarpet": [
{
"colorbar": {
"outlinewidth": 0,
"ticks": ""
},
"type": "contourcarpet"
}
],
"heatmap": [
{
"colorbar": {
"outlinewidth": 0,
"ticks": ""
},
"colorscale": [
[
0,
"#0d0887"
],
[
0.1111111111111111,
"#46039f"
],
[
0.2222222222222222,
"#7201a8"
],
[
0.3333333333333333,
"#9c179e"
],
[
0.4444444444444444,
"#bd3786"
],
[
0.5555555555555556,
"#d8576b"
],
[
0.6666666666666666,
"#ed7953"
],
[
0.7777777777777778,
"#fb9f3a"
],
[
0.8888888888888888,
"#fdca26"
],
[
1,
"#f0f921"
]
],
"type": "heatmap"
}
],
"heatmapgl": [
{
"colorbar": {
"outlinewidth": 0,
"ticks": ""
},
"colorscale": [
[
0,
"#0d0887"
],
[
0.1111111111111111,
"#46039f"
],
[
0.2222222222222222,
"#7201a8"
],
[
0.3333333333333333,
"#9c179e"
],
[
0.4444444444444444,
"#bd3786"
],
[
0.5555555555555556,
"#d8576b"
],
[
0.6666666666666666,
"#ed7953"
],
[
0.7777777777777778,
"#fb9f3a"
],
[
0.8888888888888888,
"#fdca26"
],
[
1,
"#f0f921"
]
],
"type": "heatmapgl"
}
],
"histogram": [
{
"marker": {
"pattern": {
"fillmode": "overlay",
"size": 10,
"solidity": 0.2
}
},
"type": "histogram"
}
],
"histogram2d": [
{
"colorbar": {
"outlinewidth": 0,
"ticks": ""
},
"colorscale": [
[
0,
"#0d0887"
],
[
0.1111111111111111,
"#46039f"
],
[
0.2222222222222222,
"#7201a8"
],
[
0.3333333333333333,
"#9c179e"
],
[
0.4444444444444444,
"#bd3786"
],
[
0.5555555555555556,
"#d8576b"
],
[
0.6666666666666666,
"#ed7953"
],
[
0.7777777777777778,
"#fb9f3a"
],
[
0.8888888888888888,
"#fdca26"
],
[
1,
"#f0f921"
]
],
"type": "histogram2d"
}
],
"histogram2dcontour": [
{
"colorbar": {
"outlinewidth": 0,
"ticks": ""
},
"colorscale": [
[
0,
"#0d0887"
],
[
0.1111111111111111,
"#46039f"
],
[
0.2222222222222222,
"#7201a8"
],
[
0.3333333333333333,
"#9c179e"
],
[
0.4444444444444444,
"#bd3786"
],
[
0.5555555555555556,
"#d8576b"
],
[
0.6666666666666666,
"#ed7953"
],
[
0.7777777777777778,
"#fb9f3a"
],
[
0.8888888888888888,
"#fdca26"
],
[
1,
"#f0f921"
]
],
"type": "histogram2dcontour"
}
],
"mesh3d": [
{
"colorbar": {
"outlinewidth": 0,
"ticks": ""
},
"type": "mesh3d"
}
],
"parcoords": [
{
"line": {
"colorbar": {
"outlinewidth": 0,
"ticks": ""
}
},
"type": "parcoords"
}
],
"pie": [
{
"automargin": true,
"type": "pie"
}
],
"scatter": [
{
"marker": {
"colorbar": {
"outlinewidth": 0,
"ticks": ""
}
},
"type": "scatter"
}
],
"scatter3d": [
{
"line": {
"colorbar": {
"outlinewidth": 0,
"ticks": ""
}
},
"marker": {
"colorbar": {
"outlinewidth": 0,
"ticks": ""
}
},
"type": "scatter3d"
}
],
"scattercarpet": [
{
"marker": {
"colorbar": {
"outlinewidth": 0,
"ticks": ""
}
},
"type": "scattercarpet"
}
],
"scattergeo": [
{
"marker": {
"colorbar": {
"outlinewidth": 0,
"ticks": ""
}
},
"type": "scattergeo"
}
],
"scattergl": [
{
"marker": {
"colorbar": {
"outlinewidth": 0,
"ticks": ""
}
},
"type": "scattergl"
}
],
"scattermapbox": [
{
"marker": {
"colorbar": {
"outlinewidth": 0,
"ticks": ""
}
},
"type": "scattermapbox"
}
],
"scatterpolar": [
{
"marker": {
"colorbar": {
"outlinewidth": 0,
"ticks": ""
}
},
"type": "scatterpolar"
}
],
"scatterpolargl": [
{
"marker": {
"colorbar": {
"outlinewidth": 0,
"ticks": ""
}
},
"type": "scatterpolargl"
}
],
"scatterternary": [
{
"marker": {
"colorbar": {
"outlinewidth": 0,
"ticks": ""
}
},
"type": "scatterternary"
}
],
"surface": [
{
"colorbar": {
"outlinewidth": 0,
"ticks": ""
},
"colorscale": [
[
0,
"#0d0887"
],
[
0.1111111111111111,
"#46039f"
],
[
0.2222222222222222,
"#7201a8"
],
[
0.3333333333333333,
"#9c179e"
],
[
0.4444444444444444,
"#bd3786"
],
[
0.5555555555555556,
"#d8576b"
],
[
0.6666666666666666,
"#ed7953"
],
[
0.7777777777777778,
"#fb9f3a"
],
[
0.8888888888888888,
"#fdca26"
],
[
1,
"#f0f921"
]
],
"type": "surface"
}
],
"table": [
{
"cells": {
"fill": {
"color": "#EBF0F8"
},
"line": {
"color": "white"
}
},
"header": {
"fill": {
"color": "#C8D4E3"
},
"line": {
"color": "white"
}
},
"type": "table"
}
]
},
"layout": {
"annotationdefaults": {
"arrowcolor": "#2a3f5f",
"arrowhead": 0,
"arrowwidth": 1
},
"autotypenumbers": "strict",
"coloraxis": {
"colorbar": {
"outlinewidth": 0,
"ticks": ""
}
},
"colorscale": {
"diverging": [
[
0,
"#8e0152"
],
[
0.1,
"#c51b7d"
],
[
0.2,
"#de77ae"
],
[
0.3,
"#f1b6da"
],
[
0.4,
"#fde0ef"
],
[
0.5,
"#f7f7f7"
],
[
0.6,
"#e6f5d0"
],
[
0.7,
"#b8e186"
],
[
0.8,
"#7fbc41"
],
[
0.9,
"#4d9221"
],
[
1,
"#276419"
]
],
"sequential": [
[
0,
"#0d0887"
],
[
0.1111111111111111,
"#46039f"
],
[
0.2222222222222222,
"#7201a8"
],
[
0.3333333333333333,
"#9c179e"
],
[
0.4444444444444444,
"#bd3786"
],
[
0.5555555555555556,
"#d8576b"
],
[
0.6666666666666666,
"#ed7953"
],
[
0.7777777777777778,
"#fb9f3a"
],
[
0.8888888888888888,
"#fdca26"
],
[
1,
"#f0f921"
]
],
"sequentialminus": [
[
0,
"#0d0887"
],
[
0.1111111111111111,
"#46039f"
],
[
0.2222222222222222,
"#7201a8"
],
[
0.3333333333333333,
"#9c179e"
],
[
0.4444444444444444,
"#bd3786"
],
[
0.5555555555555556,
"#d8576b"
],
[
0.6666666666666666,
"#ed7953"
],
[
0.7777777777777778,
"#fb9f3a"
],
[
0.8888888888888888,
"#fdca26"
],
[
1,
"#f0f921"
]
]
},
"colorway": [
"#636efa",
"#EF553B",
"#00cc96",
"#ab63fa",
"#FFA15A",
"#19d3f3",
"#FF6692",
"#B6E880",
"#FF97FF",
"#FECB52"
],
"font": {
"color": "#2a3f5f"
},
"geo": {
"bgcolor": "white",
"lakecolor": "white",
"landcolor": "#E5ECF6",
"showlakes": true,
"showland": true,
"subunitcolor": "white"
},
"hoverlabel": {
"align": "left"
},
"hovermode": "closest",
"mapbox": {
"style": "light"
},
"paper_bgcolor": "white",
"plot_bgcolor": "#E5ECF6",
"polar": {
"angularaxis": {
"gridcolor": "white",
"linecolor": "white",
"ticks": ""
},
"bgcolor": "#E5ECF6",
"radialaxis": {
"gridcolor": "white",
"linecolor": "white",
"ticks": ""
}
},
"scene": {
"xaxis": {
"backgroundcolor": "#E5ECF6",
"gridcolor": "white",
"gridwidth": 2,
"linecolor": "white",
"showbackground": true,
"ticks": "",
"zerolinecolor": "white"
},
"yaxis": {
"backgroundcolor": "#E5ECF6",
"gridcolor": "white",
"gridwidth": 2,
"linecolor": "white",
"showbackground": true,
"ticks": "",
"zerolinecolor": "white"
},
"zaxis": {
"backgroundcolor": "#E5ECF6",
"gridcolor": "white",
"gridwidth": 2,
"linecolor": "white",
"showbackground": true,
"ticks": "",
"zerolinecolor": "white"
}
},
"shapedefaults": {
"line": {
"color": "#2a3f5f"
}
},
"ternary": {
"aaxis": {
"gridcolor": "white",
"linecolor": "white",
"ticks": ""
},
"baxis": {
"gridcolor": "white",
"linecolor": "white",
"ticks": ""
},
"bgcolor": "#E5ECF6",
"caxis": {
"gridcolor": "white",
"linecolor": "white",
"ticks": ""
}
},
"title": {
"x": 0.05
},
"xaxis": {
"automargin": true,
"gridcolor": "white",
"linecolor": "white",
"ticks": "",
"title": {
"standoff": 15
},
"zerolinecolor": "white",
"zerolinewidth": 2
},
"yaxis": {
"automargin": true,
"gridcolor": "white",
"linecolor": "white",
"ticks": "",
"title": {
"standoff": 15
},
"zerolinecolor": "white",
"zerolinewidth": 2
}
}
},
"title": {
"text": "Loss by eliminated features"
},
"updatemenus": [
{
"active": 0,
"buttons": [
{
"args": [
{
"visible": [
true,
true,
false,
false
]
}
],
"label": "Hide features",
"method": "update"
},
{
"args": [
{
"visible": [
true,
true,
true,
false
]
}
],
"label": "Show indices",
"method": "update"
},
{
"args": [
{
"visible": [
true,
true,
false,
true
]
}
],
"label": "Show names",
"method": "update"
}
],
"pad": {
"r": 10,
"t": 10
},
"showactive": true,
"x": -0.25,
"xanchor": "left",
"y": 1.03,
"yanchor": "top"
}
],
"xaxis": {
"gridcolor": "rgb(255,255,255)",
"showgrid": true,
"showline": false,
"showticklabels": true,
"tickcolor": "rgb(127,127,127)",
"ticks": "outside",
"title": {
"text": "number of removed features"
},
"zeroline": false
},
"yaxis": {
"gridcolor": "rgb(255,255,255)",
"showgrid": true,
"showline": false,
"showticklabels": true,
"tickcolor": "rgb(127,127,127)",
"tickfont": {
"color": "rgb(51,160,44)"
},
"ticks": "outside",
"title": {
"font": {
"color": "rgb(51,160,44)"
},
"text": "loss value"
},
"zeroline": false
}
}
}
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"# Feature selection do catboost\n",
"model = CatBoostClassifier(**params)\n",
"\n",
"num_features_to_select=1\n",
"steps=len(df_train.columns)\n",
"features_for_select=list(df_train.columns.drop(\"target\"))\n",
"\n",
"print(f\"Selecting {num_features_to_select} features out of {len(features_for_select)} using {steps} steps\")\n",
"\n",
"# Criando objeto para capturar os logs do feature selection\n",
"from io import StringIO\n",
"text_results = StringIO()\n",
"\n",
"feature_selection_results = model.select_features(\n",
" pool_train,\n",
" eval_set=pool_eval,\n",
" features_for_select=features_for_select,\n",
" num_features_to_select=num_features_to_select,\n",
" steps=steps,\n",
" plot=True,\n",
" algorithm=EFeaturesSelectionAlgorithm.RecursiveByShapValues,\n",
" shap_calc_type=EShapCalcType.Regular,\n",
" train_final_model=True,\n",
" log_cout=text_results,\n",
")"
]
},
{
"cell_type": "code",
"execution_count": 215,
"id": "7bef49e1",
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"metric_ids=[0, 4, 8, 11, 14, 17, 20, 22, 24, 27, 29, 30, 32, 34, 35, 37, 38, 39, 40, 41, 42, 43, 44, 44, 45, 46, 46, 47, 47, 48, 48, 49, 49, 49, 50, 50, 50, 51, 51, 51, 51, 52, 52, 52, 52, 52, 52, 52, 53, 53, 53, 53, 53, 53]\n",
"metric_values=[0.5757575758, 0.5396825397, 0.5846153846, 0.5625, 0.6176470588, 0.676056338, 0.5172413793, 0.6060606061, 0.5901639344, 0.5517241379, 0.5625, 0.5714285714, 0.59375, 0.5454545455, 0.59375, 0.5757575758, 0.6268656716, 0.5714285714, 0.5846153846, 0.676056338, 0.6567164179, 0.6764705882, 0.6956521739, 0.6956521739, 0.5714285714, 0.6470588235, 0.6470588235, 0.6376811594, 0.6376811594, 0.5573770492, 0.5573770492, 0.5573770492, 0.5573770492, 0.5573770492, 0.6031746032, 0.6031746032, 0.6031746032, 0.5901639344, 0.5901639344, 0.5901639344, 0.5901639344, 0.2857142857, 0.2857142857, 0.2857142857, 0.2857142857, 0.2857142857, 0.2857142857, 0.2857142857, 0.2909090909, 0.2909090909, 0.2909090909, 0.2909090909, 0.2909090909, 0.2909090909]\n",
"metric_best_iters=[318, 284, 391, 245, 474, 546, 157, 410, 196, 154, 235, 290, 233, 285, 229, 269, 385, 195, 356, 485, 297, 568, 382, 382, 174, 340, 340, 342, 342, 216, 216, 292, 292, 292, 329, 329, 329, 318, 318, 318, 318, 78, 78, 78, 78, 78, 78, 78, 182, 182, 182, 182, 182, 182]\n",
"Highest metric is best.\n",
"best_metric_id=44 best_metric=0.6956521739\n"
]
}
],
"source": [
"# Tratamento dos logs para extrair os valores das melhores métricas\n",
"import re\n",
"import numpy as np\n",
"\n",
"num_results = re.findall(\"Feature #([0-9]+) eliminated|bestTest = ([0-9\\.]+)|bestIteration = ([0-9]+)\", text_results.getvalue())\n",
"\n",
"metric_ids = []\n",
"metric_values = []\n",
"metric_best_iters = []\n",
"id = 0\n",
"for feature_eliminated, best_test, best_iteration in num_results:\n",
" if feature_eliminated:\n",
" id += 1\n",
" if best_test:\n",
" metric_ids.append(id)\n",
" metric_values.append(float(best_test))\n",
" if best_iteration:\n",
" metric_best_iters.append(int(best_iteration))\n",
"print(f\"{metric_ids=}\\n{metric_values=}\\n{metric_best_iters=}\")\n",
"\n",
"metric_ids = np.array(metric_ids)\n",
"metric_values = np.array(metric_values)\n",
"\n",
"METRIC_LESSER_IS_BETTER = False\n",
"if METRIC_LESSER_IS_BETTER:\n",
" print(\"Lowest metric is best.\")\n",
" best_metric_id = int(metric_ids[metric_values.argmin()])\n",
" best_metric = float(metric_values.min())\n",
"else:\n",
" print(\"Highest metric is best.\")\n",
" best_metric_id = int(metric_ids[metric_values.argmax()])\n",
" best_metric = float(metric_values.max())\n",
"print(f\"{best_metric_id=} {best_metric=}\")"
]
},
{
"cell_type": "code",
"execution_count": 216,
"id": "1328700b",
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"len(losses)=54 losses=[0.1477282443515885, 0.14360728665267491, 0.1431814290361284, 0.14298460583655714, 0.14282510476834823, 0.138720506542803, 0.1379514449745828, 0.1373899846007138, 0.13718637249828747, 0.1396625950309679, 0.14072820600529518, 0.14097406332773885, 0.13595778276128634, 0.1346676493523341, 0.13361166554323825, 0.13187604508174725, 0.13120878267479325, 0.13104644007439964, 0.1517044460271607, 0.15171273628228008, 0.1517137533959223, 0.13722506984594268, 0.13210261129011047, 0.1443178728641258, 0.14431787286941414, 0.1458053618311109, 0.14729285079280766, 0.14878033975450441, 0.14132961172824576, 0.14132961172824576, 0.14027958275842997, 0.14027958275843014, 0.14037541750914248, 0.13926717055366758, 0.13797897441609677, 0.14212675332306965, 0.14041561086454277, 0.13757352885677757, 0.13510393460777329, 0.14409481128293467, 0.14084513882448946, 0.12712209783860534, 0.1378335582278886, 0.13478887732444192, 0.13403321273113225, 0.14991679743231506, 0.1350558793736135, 0.14042553951270548, 0.15684548881410224, 0.15089630427895054, 0.15390386308252524, 0.15754301616669614, 0.1922590970216916, 0.194306053544715]\n",
"Lowest loss is best.\n",
"best_loss_id=41 best_loss=0.12712209783860534\n"
]
}
],
"source": [
"# Extraindo melhor loss\n",
"LOSS_LESSER_IS_BETTER = True\n",
"losses = feature_selection_results[\"loss_graph\"][\"loss_values\"]\n",
"print(f\"{len(losses)=} {losses=}\")\n",
"losses = np.array(losses)\n",
"if LOSS_LESSER_IS_BETTER:\n",
" print(\"Lowest loss is best.\")\n",
" best_loss_id = int(losses.argmin())\n",
"else:\n",
" print(\"Highest loss is best.\")\n",
" best_loss_id = int(losses.argmax())\n",
"best_loss = float(losses[best_loss_id])\n",
"print(f\"{best_loss_id=} {best_loss=}\")\n"
]
},
{
"cell_type": "code",
"execution_count": 217,
"id": "ca6525b7",
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"len(features_ordered)=54 features_ordered=['elevation', 'hillshade_3pm', 'horizontal_distance_to_roadways', 'soil_type_38', 'wilderness_area_0', 'wilderness_area_3', 'horizontal_distance_to_fire_points', 'soil_type_9', 'horizontal_distance_to_hydrology', 'soil_type_37', 'soil_type_29', 'soil_type_3', 'wilderness_area_2', 'hillshade_9am', 'soil_type_30', 'soil_type_32', 'soil_type_4', 'soil_type_11', 'soil_type_1', 'soil_type_21', 'soil_type_23', 'soil_type_28', 'soil_type_22', 'soil_type_35', 'soil_type_25', 'soil_type_16', 'soil_type_18', 'soil_type_24', 'soil_type_20', 'soil_type_14', 'soil_type_7', 'soil_type_8', 'soil_type_19', 'soil_type_10', 'soil_type_6', 'soil_type_26', 'hillshade_noon', 'soil_type_17', 'soil_type_34', 'soil_type_33', 'soil_type_5', 'soil_type_27', 'soil_type_36', 'wilderness_area_1', 'soil_type_12', 'aspect', 'soil_type_15', 'soil_type_2', 'slope', 'soil_type_31', 'soil_type_0', 'soil_type_13', 'soil_type_39', 'vertical_distance_to_hydrology']\n",
"Choosing best number of features based on best metric.\n",
"best_n_features=10\n",
"features_best=['elevation', 'hillshade_3pm', 'horizontal_distance_to_roadways', 'soil_type_38', 'wilderness_area_0', 'wilderness_area_3', 'horizontal_distance_to_fire_points', 'soil_type_9', 'horizontal_distance_to_hydrology', 'soil_type_37']\n"
]
}
],
"source": [
"# Montando lista de features ordenada pela feature selection e pegando melhor subconjunto de features\n",
"features_ordered = (\n",
" feature_selection_results['selected_features_names'] + \n",
" feature_selection_results['eliminated_features_names'][::-1]\n",
")\n",
"print(f\"{len(features_ordered)=} {features_ordered=}\")\n",
"\n",
"BEST_N_FEATURES_BY_METRIC = True\n",
"if BEST_N_FEATURES_BY_METRIC:\n",
" print(\"Choosing best number of features based on best metric.\")\n",
" best_n_features = len(features_ordered) - best_metric_id\n",
"else:\n",
" print(\"Choosing best number of features based on best loss.\")\n",
" best_n_features = len(features_ordered) - best_loss_id\n",
"print(f\"{best_n_features=}\")\n",
"features_best = features_ordered[:best_n_features]\n",
"print(f\"{features_best=}\")"
]
},
{
"cell_type": "code",
"execution_count": 219,
"id": "c152cd9f",
"metadata": {},
"outputs": [
{
"data": {
"image/png": "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",
"text/plain": [
"<Figure size 640x480 with 2 Axes>"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"# Plot resumo da seleção de features\n",
"import matplotlib.pyplot as plt\n",
"import numpy as np\n",
"\n",
"fig, ax1 = plt.subplots()\n",
"losses = np.array(losses)\n",
"ax1.plot(losses, color='orange', label='Loss Values', marker='.')\n",
"ax1.scatter(metric_ids, losses[metric_ids], color='orange', marker=\"o\")\n",
"ax1.scatter(losses.argmin(), losses.min(), color=\"red\", label=f\"Min {losses.min():0.2f}\", marker=\"v\")\n",
"ax1.scatter(losses.argmax(), losses.max(), color=\"blue\", label=f\"Max {losses.max():0.2f}\", marker=\"^\")\n",
"ax1.grid()\n",
"ax1.set_xlabel(\"Num Features Removed\")\n",
"ax1.set_ylabel(\"Loss Values\", color='orange')\n",
"\n",
"ax2 = ax1.twinx()\n",
"metric_ids = np.array(metric_ids)\n",
"metric_values = np.array(metric_values)\n",
"ax2.plot(metric_ids, metric_values, color='green', label='Metric Values', marker='.')\n",
"ax2.scatter(metric_ids[metric_values.argmin()], metric_values.min(), color=\"red\", label=f\"Min {metric_values.min():0.2f}\", marker=\"v\")\n",
"ax2.scatter(metric_ids[metric_values.argmax()], metric_values.max(), color=\"blue\", label=f\"Max {metric_values.max():0.2f}\", marker=\"^\")\n",
"ax2.set_ylabel(\"Metric Values\", color='green')\n",
"\n",
"fig.legend()\n",
"plt.tight_layout()\n",
"plt.show()"
]
},
{
"cell_type": "code",
"execution_count": 220,
"id": "2c91f39d",
"metadata": {},
"outputs": [
{
"data": {
"application/vnd.jupyter.widget-view+json": {
"model_id": "a7ec5a1d71624cc7b02dfb1037f8705f",
"version_major": 2,
"version_minor": 0
},
"text/plain": [
"MetricVisualizer(layout=Layout(align_self='stretch', height='500px'))"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"Learning rate set to 0.033379\n",
"0:\tlearn: 0.1695761\ttest: 0.1363636\tbest: 0.1363636 (0)\ttotal: 7.46ms\tremaining: 14.9s\n",
"100:\tlearn: 0.4505263\ttest: 0.4074074\tbest: 0.4074074 (92)\ttotal: 691ms\tremaining: 13s\n",
"200:\tlearn: 0.6068702\ttest: 0.5483871\tbest: 0.5483871 (186)\ttotal: 1.4s\tremaining: 12.5s\n",
"300:\tlearn: 0.6881720\ttest: 0.6363636\tbest: 0.6363636 (266)\ttotal: 2.08s\tremaining: 11.8s\n",
"400:\tlearn: 0.7547170\ttest: 0.6567164\tbest: 0.6764706 (330)\ttotal: 2.94s\tremaining: 11.7s\n",
"Stopped by overfitting detector (100 iterations wait)\n",
"\n",
"bestTest = 0.6764705882\n",
"bestIteration = 330\n",
"\n",
"Shrink model to first 331 iterations.\n"
]
},
{
"data": {
"text/plain": [
"<catboost.core.CatBoostClassifier at 0x1a4dabb9b90>"
]
},
"execution_count": 220,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"# Treinando modelo com as melhores features\n",
"pool_train = Pool(data=df_train[features_best], label=df_train[\"target\"])\n",
"pool_eval = Pool(data=df_eval[features_best], label=df_eval[\"target\"])\n",
"pool_test = Pool(data=df_test[features_best], label=df_test[\"target\"])\n",
"\n",
"model_final = CatBoostClassifier(**params)\n",
"\n",
"model_final.fit(\n",
" pool_train,\n",
" eval_set=pool_eval,\n",
" plot=True\n",
")"
]
},
{
"cell_type": "code",
"execution_count": 222,
"id": "84d0a81e",
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
" precision recall f1-score support\n",
"\n",
" 0 0.95 0.99 0.97 1062\n",
" 1 0.77 0.41 0.54 100\n",
"\n",
" accuracy 0.94 1162\n",
" macro avg 0.86 0.70 0.75 1162\n",
"weighted avg 0.93 0.94 0.93 1162\n",
"\n",
"AUC: 0.9431073446327685\n"
]
}
],
"source": [
"# Classification report\n",
"import numpy as np\n",
"from sklearn.metrics import classification_report, roc_auc_score\n",
"\n",
"y_pred_proba = model_final.predict_proba(pool_test)\n",
"y_pred = model_final.predict(pool_test)\n",
"\n",
"print(classification_report(df_test[\"target\"], y_pred, zero_division=0))\n",
"print(f\"AUC: {roc_auc_score(df_test['target'], y_pred_proba[:, 1], multi_class='ovr')}\")"
]
},
{
"cell_type": "code",
"execution_count": 224,
"id": "9256078e",
"metadata": {},
"outputs": [
{
"data": {
"image/png": "iVBORw0KGgoAAAANSUhEUgAAAjcAAAGwCAYAAABVdURTAAAAOnRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjEwLjgsIGh0dHBzOi8vbWF0cGxvdGxpYi5vcmcvwVt1zgAAAAlwSFlzAAAPYQAAD2EBqD+naQAATs5JREFUeJzt3Qd4lFXWwPEDKZAEAqGH3nuJgLA0kRqlhSpWsAAioi7s2gtgw7WgfiuKioiiLkoVpAUpCgLSXaVL70U6gRDCfM+57oxJCCETpr7v//c8Q/IOk8mbm8nMmXvPOTeXw+FwCAAAgEXk9vcJAAAAeBLBDQAAsBSCGwAAYCkENwAAwFIIbgAAgKUQ3AAAAEshuAEAAJYSKjZz+fJlOXDggOTPn19y5crl79MBAADZoG35zpw5IyVLlpTcubOem7FdcKOBTZkyZfx9GgAAIAf27t0rpUuXzvI2tgtudMbGOTjR0dEeve+UlBRJTEyU9u3bS1hYmEfvG4yzr/F4Zpythsd0cI/z6dOnzeSE83U8K7YLbpxLURrYeCO4iYyMNPdLcOM9jLNvMM6Ms9XwmLbGOGcnpYSEYgAAYCkENwAAwFIIbgAAgKUQ3AAAAEshuAEAAJZCcAMAACyF4AYAAFgKwQ0AALAUghsAAGApBDcAAMBS/Brc/Pjjj9K5c2ezw6e2U54+ffo1v2bx4sVSv359yZMnj1SuXFnGjx/vk3MFAADBwa/Bzblz56RevXoyevTobN1+586d0rFjR2nVqpWsX79e/v73v0u/fv1k3rx5Xj9XAAAQHPy6ceatt95qLtk1ZswYqVChgrz11lvmuEaNGrJ06VJ5++23JT4+3otnCsCqHA6HnE9J9dv3T0m5JMmpIkkXL0mY49obAoKxDnQp/3tM69+WvwTVruDLly+Xtm3bprtOgxqdwbma5ORkc0m7Zbpz11K9eJLz/jx9v2Cc/cEOj2d98r197CpZu+ekn88kVJ5YudDP52AXjLVvhErr1slSIBs7eGeXO89FQRXcHDp0SIoXL57uOj3WgOX8+fMSERFxxdeMHDlSRowYccX1iYmJZkt2b5g/f75X7heMsz9Y+fGs7y7X7gmqp0EgIOWRFNEw5oKEua5buHCh5Anx3PdISkrK9m0t/1f99NNPy9ChQ13HGgiVKVNG2rdvL9HR0R79XhpV6gtBu3btJCzsr18wPMtO4+zPJROdWtYnp9atW0tYmDWfKs5fTBVZ+YP5fMWTLSUi3IPPxNlkh3EOFIy1d+zbu1fmzJophQoVlm49e0lq6mXzmO4Y31bCw8M99n2cKy/ZEVR/SSVKlJDDhw+nu06PNUjJbNZGaVWVXjLSF0VvvTB6875hn3HWwKbnmOWyZvcJP55FqMjKH8UOoqPySmR4qF+CdX13WyAqr6Ufz4GAsfb8c9SSJUtMFbN+HpE3r4TKZckXldc8pjWw8eRj2p37CqrgpkmTJjJ79ux01+k7eL0esBqdsfFvYGMfDcvFSESY72dtgGB19uxZmTZtmuzYscMca+Vzhw4dTEATCHl6of4enN9//z1dqbeWeBcqVEjKli1rlpT2798vn3/+ufn/gQMHynvvvSdPPPGE3H///Wba65tvvpFZs2b58acAvG/1c20l0sdLJvoENW9eosTHt7f8jIIGNtprC8C16Wv11KlTzWu4PjdoUBMXFyeBxK/BzerVq03PGidnbkzfvn1Nc76DBw/Knj17XP+vZeAayAwZMkTeffddKV26tIwdO5YycFhS2ipKDWx8vWSSksthppb1+5ILAkBdvnzZrKBoYFO0aFHp1auX+Rho/Brc3HzzzVnWwWfWfVi/Zt26dV4+M8C/9O+i15jl/BoABJTcuXNLjx49zOSEtmIJ1Fld9pYCAjTfZuPBPysDasZGkw8CwG+2b98ua9asSVfc06lTp4ANbIIuoRiwS1l2kpYo/8+kgU3IBwHglyWoRYsWmZ0AdMZG94GMjY0Nit8EwQ0Q4GXZ5LkC8LXTp0/LlClTXHmvN9xwQ0Dm1lwNwQ0QwGXZlCgD8LVt27aZMm/t/K+l3V26dJFatWoF1S+C4Aa2lt2lprTLRL4sy6ZEGYAvLViwwCxDKV2C6tmzp2nPEmwIbmBbOV1q8kdZNgD4QsT/uv03atTIbHETGhqcz3XBedaAn5aaWCYCYDUXL1507QGlHf+1h5w20g1mBDeAG0tNLBMBsIrU1FSzhZGWevfv398EONqpO9gDG0VwA9vm2KTNo2GpCYCdnDhxQiZPniwHDhwwx1u2bJE6deqIVRDcwDYCY5dtAPCvjRs3yowZMyQ5OVny5s0rXbt2lWrVqlnq10JwA7F7jg15NADs4NKlS5KYmCirVq0yx2XKlDFbKRQoUECshuAGtpF2G7O0OTbk0QCwg8Q0gU2zZs3MxtUhIb5pa+FrBDew5UaU5NgAsJubbrpJdu/ebUq8K1euLFbGxpmwBTaiBGA3KSkp8uuvv7qO8+XLJwMHDrR8YKOYuYEtOg6zESUAOzl27JhMmjRJjhw5Yja9dG6foKXedkBwA9tVQ9nkbxuATf3yyy8ya9YsM3MTFRXl6jpsJwQ3sFXHYSqjAFi50/CcOXNk/fr15rhChQrSrVs3yZ8/v9gNwQ1s1XGYyigAVqTLT9qU7+jRo2bpqWXLltKiRQuzJGVHBDewZKk31VAA7NZxWAMbTRru0aOHlC9fXuyM4AaWLPUGADs87zkThLXDcOfOnc3HqKgosTt7zlfBcij1BmAnhw4dkk8//VROnTrluq5+/foENv/DzA2CrrQ7JeWSJKdqefclCXP8+a6FUm8AdnlOXLNmjcydO9fs6q1dh3v16uXv0wo4BDcI0tLuUHli5cJMv45SbwBWpBtdzpw5UzZs2GCOq1SpIh07dvT3aQUkghsEdWl3RpR6A7CigwcPmmqo48ePmwqoNm3aSJMmTWzTlM9dBDcIutJubUw1b16ixMe3l7CwsHS3o9QbgNXs3LlTvvzyS7MMpTt49+zZU0qXLu3v0wpoBDcI2CWptHk0aUu7U3I5JE+IXhcqYWE8hAFYmwYyhQsXlpiYGElISLBlx2F38cqAoN1GAQCs3JSvSJEiZglKZ6j79u1rghqWobKHUnAEfK4NeTQA7PTmbvny5fLhhx/K0qVLXddHRkYS2LiBmRsEXMl32uUozbUpHBXOHzUAyzt//rxMnz5dtm7d6pq9SduoD9lHcIOAXobSXBv+sAFY3d69e0011OnTpyUkJETi4+OlYcOGPP/lEMENArbkm+UoAHZ4c7ds2TJZsGCB+bxQoUKmGio2NtbfpxbUCG7gk67CWcm4DOUs+aasG4DVad+aRYsWmefN2rVrS6dOnSRPnjz+Pq2gR3CDgKp0YjdvAHaiJd4dOnQwz526NxTL8J5BcAO/dxV2YhkKgNVpEKNVUBUrVpRSpUqZ6zSogWcR3MBr0i4xZQfLUACs7OzZszJt2jTZsWOHrF27Vh566CEJDw/392lZEsENPJpXc7WuwgBg9y0Upk6dagKc0NBQadmyJYGNF/HKA7fRQRgAsufy5cvy448/yg8//GCOixYtKr169TIf4T0EN/BKXg35MwDsLjk5WSZOnCi7du0yx3FxcSZ5OOOGv/A8ghtcdwfhzPJqyJ8BYHeaT6OBjF60xLtu3br+PiXbILjBdXcQJq8GAP5ahkpNTTUBjZZ1d+3aVZKSkswmmPAdNs7ENdFBGACuTbdO+Oyzz2TWrFnpNrwksPE9Zm5wTQ7HX5/TQRgArrRt2zZT5q2bXx46dEhuvvlmKViwIEPlJwQ3uOaSVK8xy13HLEMBwF90CWrhwoVmfyile0Lp3lAENv5FcINrLkltPHjafF4zNtokCgMARE6dOmV28t63b58ZjkaNGkm7du1MHxv4F78BZNukgU3Y9wQA/jer/cUXX8ixY8fMRpcJCQlSo0YNxiZAENwgy9LvtGXfuXIxWADw5/NhLrnllltk8eLF0r17d4mJiWFgAgjBDa5AB2IAuNKJEyfk+PHjUqlSJXOsH3UDTHbyDjwEN8hW6TcdhwHY2caNG2XGjBnm8wEDBkihQoXM5wQ2gYngxsays/mls/SbjsMA7OjSpUuSmJgoq1atMselS5eWkBAKKwIdwY1NZXfpidJvAHb1xx9/mGoo7VujmjZtKq1btya4CQIENzbF5pcAcHW//fabzJw5Uy5evCgRERHSrVs3qVKlCkMWJAhuwOaXAJCB9q7RwKZs2bLSo0cPiY6OZoyCCMENWHoCgP8t1zsThLUZnyYNN2zYUHLnZhvGYMNvDABge//973/lq6++Mrt6K00a1o7DBDbBiZkbAIBt6dLTnDlzZP369eZ43bp10qBBA3+fFq4TwY1NS7/TlnsDgB0dOXLEVEMdPXrUHLds2VJuuOEGf58WPIDgxiboOgwAfz0f6gzN7NmzTR+bfPnymS0UKlSowBBZhN9zbkaPHi3ly5eXvHnzSuPGjWXlypVZ3v6dd96RatWqmdK8MmXKyJAhQ+TChQs+O1+rlX7TeRiA3SxZssR0G9bARrdPGDhwIIGNxfh15ubrr7+WoUOHypgxY0xgo4FLfHy8bNmyRYoVK3bF7TXZ66mnnpJx48aZZkpbt26Ve++912S3jxo1yi8/QzBugOnsOqzoPAzAbmrWrGk6Djdr1kyaN2/OFgoW5NfgRgOS/v37y3333WeONciZNWuWCV40iMlo2bJl5sF45513mmOd8bnjjjvk559/vur3SE5ONhen06dPm48pKSnm4knO+/P0/V5PYHP72FWyds/JdNeH5bosYbn+nLTTdy7BJtDG2aoYZ8bZKvS5UPNrnPtBFShQQAYNGmRWAILxOdCuzx0pbtxfLof+1v2UoR4ZGWmSubp27eq6vm/fvnLy5En59ttvM5250Qek7vOhJXo7duyQjh07yj333CPPPPNMpt9n+PDhMmLEiEzvS7+/lSWnijyxMn38WiG/Qx6rlSr/a+UAAJaWmpoqe/fuNa8rlStXNvk1CE5JSUlmcuPUqVPXbKrot5mbY8eOmQdd8eLF012vx5s3b870a/SH0q/TaUSNyTTi1rXSqwU26umnnzZLX2lnbjRXp3379h7vOKlR5fz5803zp7CwMPF3RdR5XYpa+YP5fMWTLSXCIhtgBsI42wHjzDgHO90Tatq0aSaw0ec9za/RGRyeO4LzucO58mK5aqnFixfLq6++Ku+//77J0fn999/lsccek5deekmef/75TL8mT5485pKRDri3Xhi9ed85rYiKjsorkeFB9esO2HG2G8aZcQ42+lyoOTU6y69vonUZSrdQKFGihKmQ4jHtG54eZ3fuy2+vdkWKFDEdIA8fPpzuej3WB2BmNIDRJah+/fqZ4zp16si5c+dkwIAB8uyzz9q+kyQVUQDsTqtntRJq06ZN5liraxMSEkx+DXl69uG34CY8PNx0gVywYIEr50bbXuvx4MGDr7relrEVtgZIyk+pQwGLiigAdqRpDRrY6GuFLovoLH+wL8XDfX5dp9BcGE0g1o3JNEFYS8F1JsZZPdWnTx8pVaqUjBw50hx37tzZVFhpB0nnspTO5uj1ziDHztLGd1rqbbVlKAC4lnr16pkVgNq1a5vXD9iTX1/9evfubdpev/DCCybxKy4uTubOnetKMt6zZ0+6mZrnnnvOROD6cf/+/VK0aFET2Lzyyitidzpz1WvMcn+fBgD41Pnz52XhwoXSpk0b0wxWXyO0Xxrsze9v7XUJ6mrLUJpAnFZoaKgMGzbMXHBlvs3Gg39mkteMjTZVUQBgZVriPWXKFFMarP3MdAsFICCCG3hm1iZtB+JJA5uwxgzA0s952tRVZ2w0VzMmJkaaNGni79NCACG4CXKZlX+TOwfAqrSwZPr06bJt2zZzXKtWLZOekFnLD9gXwY3Fyr/ZCBOAVWlupnaXP3PmjCkiufXWW6V+/frMVOMKBDcWK/8uHBXOHzoAS3J2lS9cuLD06tXrig73gBPBjYVo+Tf9HABYiSYKO5ecdD/Au+++WwoWLGh6pQFXk74jHgAAAWLnzp3y3nvvyfr1613XFStWjMAG10RwAwAIKFoBpa1AJkyYIGfPnjX7RNGFHu5gWcpCJeAAEOw0WVh38tZZG6XNXTVxmCV3uIPgJkhltQM4AASj7du3m8BGt+HRHaA7duxotlMA3EVwE6QoAQdgJSdOnJAvv/zSvHHTvBqthipSpIi/TwtBiuDGAptkUgIOINhpl+FmzZqZvaJ0byiduQFyiuDGAptkUgIOIBhpl2GdndHARrVu3ZrcGngE1VJBiE0yAQSz1NRUmT9/vuk2PHnyZHOsSBqGpzBzE+TYJBNAMNEdvDWg2bdvnzkuVaoUZd7wOIKbIM+3YZNMAMFiy5YtZtPLCxcumK7DXbp0kZo1a/r7tGBBBDdBnm8DAIFOl52+//57WbFihTkuWbKk9OzZ05VrA3gawU2QId8GQDC+Kdu9e7f5vHHjxtKuXTuzqzfgLQQ3QbwkRb4NgEAPajRJODQ01PStOXz4sFSvXt3fpwUbILgJ4iUp8m0ABKJLly5JYmKi5M2b15R3K12CYhkKvkJwE0RYkgIQ6I4fP26qoQ4ePGhmbXRvqEKFCvn7tGAzBDdBiiUpAIFmw4YNMmPGDLl48aJERERI165dCWzgFwQ3QYolKQCBIiUlRebNmydr1qwxx2XLlpUePXpIdHS0v08NNkVwAwC4rlzACRMmyN69e81x8+bNpVWrVpI7Nw3w4T8ENwCAHNO8mvr168sff/wh3bt3l0qVKjGa8DuCmyB5Z6TJxEkX/9x/BQD8vQx18uRJKVq0qDnWpOFq1aqZPBsgEBDcBEFg03PMclmz+4S/TwUA5OjRozJp0iRJTk6WBx98UCIjI82oENggkBDcBDidsckY2DQsFyMRYXT3BOBb69evl1mzZpk+Nvny5TOzN87gBggkBDdB1JF49XNtJTI8xAQ2us4NAL6gpd2zZ8+WX375xRxXrFhRunXrZgIcIBAR3ARRR2INbCLD+ZUB8B3dMkGb8h07dsy8qbr55pulRYsWvMFCQOOVMoDRkRiAv/30008msMmfP7/pXVOuXDl/nxJwTQQ3QYKOxAD8oUOHDmbjyzZt2khUVBS/BAQFuiwFCVJsAPiC7gmlm17qsrjSzS+7dOlCYIOgwswNAMAEM6tXrzbbKKSmppoeNjfccAMjA/sFNxcuXDBRPQAgeOlz+cyZM2Xjxo3muGrVqlK9enV/nxbgu2Wpy5cvy0svvSSlSpUyZYA7duww1z///PPyySef5PxMAAA+t3//fvnwww9NYKP7QbVv315uv/12mvLBXsHNyy+/LOPHj5fXX39dwsPDXdfXrl1bxo4d6+nzAwB4ybp162TcuHGmGV/BggXl/vvvlyZNmlDmDfsFN59//rl89NFHctddd0lIyF9dcuvVqyebN2/29PkBALykUKFCJtemRo0aZisFnZEHbJlzo1OYlStXznS5SjdTAwAErrS5ktqzpl+/fhIbG8tsDew9c1OzZk1ZsmTJFddrB0sy6wEgMOkMzbJly+Tdd981TfmcSpYsSWADy3F75uaFF16Qvn37mhkcna2ZOnWqbNmyxSxXfffdd945SwBAjiUlJcn06dNl27Zt5lj3iNKmfIBVuT1zk5CQYEoGv//+e9PUSYOdTZs2mevatWvnnbMEAOTInj17TDWUBjaaJ9mxY0dp3bo1owlLy1GfG900bf78+Z4/GwCAx5ahli5dKosWLTKfFy5cWHr27CklSpRghGF5bs/c6Fb3f/zxxxXXaymh/h8AwP/Wr18vCxcuNIFN3bp1ZcCAAQQ2sA23Z2527dplWnNnlJycbPJwAAD+p+05fvvtN9ODLC4ujqRh2Eq2g5sZM2a4Pte9RwoUKOA61mBnwYIFUr58ec+foQ3pO63zKamSdPHKIBIAMqMFHtqUTwMZza3RbsN33303QQ1sKdvBTdeuXc3HXLlymWqptMLCwkxg89Zbb3n+DG0Y2PQcs1zW7D7h71MBECTOnj1rKld37txpyrzj4+Ndz9eAHYW6865AVahQQVatWiVFihTx5nnZls7YZAxsGpaLkYiwv7pBA4CT7u+ngc25c+fMG00ShoEc5NzoOwP4xurn2kpkeIgJbHgHBiDjG87Fixe7mqoWK1ZMevXqxRtPIKel4PoO4YcffjD9Ey5evJju/x599FEG1kM0sIkMz9GvCICFnT592szW7N692xzXr19fbrnlFjNzAyAHwY0mrHXo0MF0vNQgRzde0zXeyMhI886B4AYAvOvSpUty8OBBCQ8Pl06dOkmdOnUYcuB6gpshQ4ZI586dZcyYMaZiasWKFebdgmblP/bYY+7eHQAgm8UGzuVpfVOpS1AxMTGmOR+A62zip42h/vGPf5gyQy031P42ZcqUkddff12eeeYZd+8OaZ64ki5eovwbwBVOnTol48ePN8nDTpUrVyawATw1c6OzNBrYKF2G0rybGjVqmFmcvXv3unt3oPwbQBZ0Y+Jvv/1Wzp8/L7Nnz5ZBgwa5noMBeCi4ueGGG0wpeJUqVaRly5Zm40zNuZkwYYLphAn3Uf4NICNtjqobFOvSvypZsqTZG4rABvBCcPPqq6/KmTNnzOevvPKK9OnTRx566CET7HzyySfu3h0yoPwbgO7VN3nyZNeWNo0bN5a2bdtKaCjVk0B2uP2X0rBhQ9fnuiw1d+5cd+8CWaD8G7A3za/58MMP5cKFC5I3b15JSEiQ6tWr+/u0gKDisYXbtWvXmpJEd40ePdps3aB/xPruZOXKldd8R/Pwww9LbGys5MmTR6pWrWrWoQHACqKjo83zWunSpeXBBx8ksAG8PXOjG2bOnz/f9Fbo16+fVKxYUTZv3ixPPfWUzJw507WfSXZ9/fXXMnToUFNWroHNO++8Y+5DE+h0VigjbRjYrl078386ZVuqVCnTxKpgwYIS3FVSbJAJ2JlWnWrvMC3M0HJvfaPorEgF4MXgRvNp+vfvb/ornDhxQsaOHSujRo2SRx55RHr37i2//fabqZpyh3693ud9991njjXImTVrlowbN84ETBnp9cePH5dly5a5OnFeaydyfdLQS9rOniolJcVcPMl5f9m9Xw1sbh+7StbuOZnuPlJyOTx6Xlbj7jiDcQ5kv/76q3lDp2/ebrvtNlcvG91ewbmnHzyD547gHmd37i+XQ19hs6Fu3bpyzz33yOOPPy5TpkwxDaT+9re/yTfffGOmT92lf8ja1VhnYJw7jivdcVyXnrT0MSPtjKzBlX6d/n/RokXlzjvvlCeffPKq73CGDx8uI0aMuOL6r776ytyPPyWnijyx8q/4skJ+hzxWK1XYyBewPg1cNGH4jz/+MMdRUVFmNpzZGiBzOrupr/mal6bLtx6Zudm+fbsJaFT37t1N1v4bb7yRo8BGafm4ljoWL1483fV6rEtdmdEGVgsXLpS77rrL5Nn8/vvvpueDRnPDhg3L9Guefvpps/SVduZGmw62b9/+moPjLj0PXbbTpbPs7PGiTfueWLnQfL7iyZZSKCqcDTK9MM7IGcbZezSgmTZtmiuw0aV2ffOoeYTwHh7TwT3OzpWX7Mh2cKMNpJwzHTptqn+EmtTr63c6+iTw0UcfmXc3DRo0MO98NMi6WnCj55nZE4YOuLdeGLN732GOP6efVXRUXglnk0yvjDOuD+PsWf/973/lu+++My8A+pzapUsX84ZOn6d4PPsGj+ngHGd37suthGLNs8mXL59r4zZtB16kSJF0t8nuxpn6dRqgHD58ON31elyiRIlMv0aDKf3h0k7bap7PoUOHzDKXJjoDQKDSgGbRokXmo+YL6iy4VopebbYaQM5kO7gpW7asfPzxx65jDUC0K3FaOqOT3eBGAxGdeVmwYIEr50ZnZvR48ODBmX5Ns2bNTK6M3s7ZpXPr1q0m6CGwARDo9M2Zdhnetm2b3HTTTeZ5jMR4wI/Bza5duzz+zTUXRhOItTFgo0aNTCn4uXPnXNVT2v1Yy71HjhxpjrUT8nvvvWd2H9cqLX2C0I7J2Q2oAMDXdLNhrdvQrWuUPqfpBYD3+LWXt5aQHz161OxPpUtLcXFxpuOxM8lYN+VMu4+KJgJrr50hQ4aY6i19gtBAR6ulACCQ6FK5Fj788ssvZildZ78LFy7s79MCbMHvG5XoEtTVlqEWL158xXVNmjRxbSQHAIFIcwe1zYVWhepyvS5BxcTE+Pu0ANvwe3ADAFahy0/r1q2TOXPmmKKL/Pnzm6ThazUbBeBZBDcA4KHAZvr06abUW1WuXNkUS2hzPgC+RXADAB6gy0/aQV0/tm7d2lR3OrdSABAEwY12K/7000/Nx3fffdc01tNpWE2Yq1WrlufPEgACdLbmwoULEhERYY5btGgh1apVu2qvLgC+8VcpUjb98MMPUqdOHfn5559l6tSpcvbsWXO9VgRcrUswAFiNBjWaNPzZZ5+5etVodSeBDRCEwY3u1v3yyy+bfSPSNs7TaViqmADYwYEDB8w2MBs3bjTtLPbu3evvUwJwPctSv/76q+kSnJEuTWnZIwBYeRlq5cqVkpiYaDqlFyhQwHQczukGwgACJLgpWLCgHDx4UCpUqJDuei1/pOumexwOd0cfgL/o5sEzZsxw7QNVvXp1s+mlM98GQBAvS91+++2mI7B2FNZKAH338tNPP8k///lPs10Csv8OsNeY5QwXECS027AGNtpt+JZbbpHbbruNwAawysyN7uX08MMPm60QUlNTpWbNmubjnXfeKc8995x3ztKCzqekysaDp83nNWOjJSLsr53OAQSetm3byvHjx6Vjx45SsmRJf58OAE8GN5pErLuDP//88/Lbb7+ZaindEK5KlSru3hX+Z9LAJvTDAAJMUlKSbN261ex5pzS/pl+/fvytAlYMbpYuXSrNmzc3PW30gutHny8gsOimvVOmTJHTp0+bpSftXaNoygdYNOdGS741mfiZZ54xZZAAYKVcOH0DN378eBPYaMdhnbEBYPHgRvs7/OMf/zDN/GrXrm2mbN944w3Zt2+fd84QAHzg3Llz8uWXX8qCBQtMkKPNSgcMGEBTPsAOwU2RIkVk8ODBpkJKt1/o1auX6dCpu97qrA4ABJtdu3bJmDFjzHNaaGiodO7cWbp16yZ58uTx96kB8PXGmbo8pR2L69WrZxKMdTYHAIKNFkboRd+86Rs2bUoKwIbBjc7c6BSu7q2ie6wkJCTIyJEjPXt2AOAluvTkTBDWJXZtaVGjRo1028oAsMmy1NNPP21mbHQJSisKdFdwbeg3YcIE09gKAALdjh07zN5Qzo1/lc5AE9gANp25+fHHH+Xxxx833Tl1ChcAgoV2VNflc30eU4sXL5ZOnTr5+7QA+Du40eUoAAg2Z86cMb1rdu/ebY61+Wh8fLy/TwuAv4Ib3Szu1ltvlbCwMPN5VnQjOQAIJL///rtMmzbNdB3WpSedrdFSbwA2Dm66du1q8mq0gkA/vxpNztOkPFw7kTHpIuME+MKGDRtM4YMqXry4qYYqXLgwgw/YPbjRderMPkfOApueY5bLmt0nGD7ABypXrmyCGS2E0GUo7WMDwNrcrpb6/PPPJTk5+YrrL168aP4P194NPG1g07BcDDuCAx6mHdP1jYTSRnz9+/c3u3kT2AD24HZwc99998mpU6cyTdbT/0P2rX6uLTuCAx6ky+KJiYnyySefyIoVK1zX02kYsJfQ62l8lfGdEhvMuScyPIRdhgEPOXnypMmt2b9/v+sNFwB7ynZwo2WTGtTopU2bNummd/Xd0s6dO2niB8AvNm/eLN9++63plp43b17TMb169er8NgCbynZw46ySWr9+vUnKy5cvn+v/tLRSN87s0aOHd84SADJx6dIlmT9/vqxcudIclypVSnr27CkFCxZkvAAby3ZwM2zYMPNRg5jevXubd0cA4E9Hjx6V1atXm8+bNGliZpVDQkL4pQA253bOTd++fb1zJgDgptjYWNNgNDo6WqpWrcr4Ach+cFOoUCHZunWr2UsqJiYmyyTY48ePZ+cuASDHy1D169c3DflUw4YNGUkA7gc3b7/9tuTPn9/1eVbBDa6OzsRAzv3xxx8yadIkOXz4sNnV+6GHHpLcud3uZgHABkLdXYq69957vXk+lkVnYiDnfv31V/nuu+9Ms9DIyEhT1EBgA8BjOTdr1641G2g6N53T8stPP/1UatasKcOHDzeVU7gSnYkB96WkpMicOXNk3bp15rhcuXKmKtM5kwwAmXF7TvfBBx80+TdKp4a1ckrfSel08RNPPOHu3dkSnYmBazt79qyMHTvWFdjcdNNN0qdPHwIbAJ4PbjSwiYuLM59rQNOyZUv56quvZPz48TJlyhR3786W6EwMZOPvJDJSoqKizOWee+6RVq1asRQFwHvbLzh3Bv/++++lU6dO5vMyZcrIsWPH3L07AHDRnBrNpdEO6Pqxe/fu5vq0TUMBwOMzN1p2+fLLL8uECRPkhx9+MDvtKt1+wVmaCQDuOnLkiHz88ccyd+5c13Ua1BDYAPD6zM0777wjd911l0yfPl2effZZqVy5srleN6xr2rSp2ycAwN50NljzajRxWPvYJCcnS1JSklmWAgCfBDd169Y1ZZkZvfHGG7Q9B+AWDWRmzZrlek6pVKmSdOvWjcAGgG+DG6c1a9bIpk2bzOdaBq4dQwEguw4dOmRmfLU5nzYGbd26tTRr1owmoQB8H9zouriWf2u+jXPn3ZMnT5pKhokTJ0rRokWv/6wsyOHw9xkAgUOXn7TK8syZM2ZfKO1dU7ZsWX+fFgC7JhQ/8sgjpv/Ehg0bzD5Sevntt9/k9OnT8uijj3rnLC2QU9BrzHJ/nwYQMLQaSosRdLNL7Z1FYAPArzM3WsmgJeA1atRwXafLUqNHj5b27dt79OSs1J1448HT5vOasdESERbi71MCfO7AgQNy4cIFqVixojmuVq2aCW7Yqw6A34Mb7XGj2y9kpNc5+9/g6iYNbMKTOWw3c7ly5Uqzm7duz6IzNQUKFDD/R2ADICCWpTTp77HHHjPvwpz2798vQ4YMkTZt2nj6/CyHDdVhJ+fPn5dvvvnGzPimpqaavaHYfw5AwM3cvPfee9KlSxcpX7686Uqs9u7dK7Vr15YvvvjCG+cIIAjt27fPbMmiBQchISHSrl07adSoEbM1AAIvuNGARncGX7BggasUXPNv2rZt643zswQqpWC3ZagVK1aY3Dxdqo6JiZGePXtKyZIl/X1qAGzCreDm66+/lhkzZpj9X3QJSiunkDUqpWA3mkej+8xpYKPFBp07d5a8efP6+7QA2Ei2g5sPPvhAHn74YalSpYpERETI1KlTZfv27aYzMa6OSinYKZB3JgjfcsstJr+mTp06LEMBCNyEYs21GTZsmGzZskXWr18vn332mbz//vvePTuLoVIKVg1qli5dapry6efO6kndqoVqKAABHdzs2LFD+vbt6zq+8847TZfRgwcPeuvcLIdKKVjNuXPnTFCjOXi///67bN682d+nBADZX5bSDe6ioqJcx7lz5zYlnVrqCcB+du/ebaqhdAsF7Th86623SvXq1f19WgDgXkLx888/n263Xk0sfuWVV1wNudSoUaMYVsDCNFFYl6EWL15slqGKFCkivXr1kmLFivn71ADAveDmpptuMvk2aTVt2tQsVzmxvp6ePvEnXUzN7hADQWHWrFmmHYSKi4szMzY05gMQlMGNvkuDe4FNzzHLZc3uEwwbLOXGG280Pa7i4+OlXr16/j4dALj+7Re8QTfd1I7H2gujcePGZh+a7Jg4caKZLeratasEYgl42sCmYbkYNsxE0C5DaRdypxIlSsjf//53AhsAAcvvwY02Bhw6dKgpM9epbn0nqO8Ijxw5kuXX7dq1S/75z39KixYtJNCtfq4tZeAISikpKaYaavz48WYPOSeWoQAEMr8HN5qA3L9/f7nvvvtMN9MxY8aYpOVx48Zd9Wt0A7677rpLRowYIRUrVpRAFxkeQj4Sgo7m02me3Z49e0w1lFZFAYAl95byJK22WrNmjTz99NPpSsx1n6rly5df9etefPFFU5nxwAMPyJIlS65Zwq4Xp9OnT7vekerFk5z3Z+7b8WenVtdxrj+bm8Gz4wzvLEP98MMPrr/BokWLSvfu3aVw4cKMuRfwePYdxjq4x9md+/NrcKP7z+gsTPHixdNdr8dXawamJaiffPKJ6ZKcHSNHjjQzPBklJiamK2v3pPnz50ty6l/DO29eouQJ8cq3sjUdZ3j+DYf2r9HmfErLvGNjY+Xnn39mqL2Mx7PvMNbBOc5JSUneDW50tuTDDz80e0tNnjxZSpUqJRMmTJAKFSpI8+bNxVt0Wvyee+6Rjz/+2DzpZofOCmlOT9qZG93ZvH379hIdHe3xqFJ/me3atZOLl3PJEysXmuvj49tLZLhf40hLSTvO2uYfnqPJ/Bs3bpQ8efKY3DddkmKcvYvHs+8w1sE9zs6Vl+xw+xVXO5JqgKE5L+vWrXMt+Zw6dUpeffVVmT17drbvSwOUkJAQOXz4cLrr9VgrMjLSYEoTiXWX4bRT6OYHCQ01+QGVKlVK9zX6JK2XjHTAvfXCqOfS84OfM3wvghtP8+bv0K60d5W+O2rQoIHkz5/fBDeMs28wzr7DWAfnOLtzX24nFL/88ssm6VdnT9J+o2bNmrkae2WXVlzok6juS5M2WNHjJk2aXHF7be3+66+/miUp56VLly7SqlUr87nOyAQCdgJHsDh58qRMmzbNLEcpba2g77YKFSrk71MDgBxzezpBZ0e0W3FGugWDPlG6S5eMdEPOhg0bSqNGjeSdd94x6/1aPaX69Oljlr00d0b74NSuXTvd1xcsWNB8zHh9oGAncAQqzWv79ttv5cKFC+aNRseOHf19SgDgn+BGl4t0919tupcx0TcnZdm9e/eWo0ePygsvvCCHDh0y7dznzp3rSjLWaXGtoApW7ASOQKNJ/Loe7kwS1jcPOvMKALYNbrQnzWOPPWb60OgU9oEDB0zJqDbU0401c2Lw4MHmkpNtH7S5GIDsOXHihCkC0L9bpcu/bdq0MblvAGDb4Oapp54yeTH6hKiJh7pEpQm7Gtw88sgj3jlLANdNk/F1yxItAoiIiDDbllStWpWRBWA5bgc3Olvz7LPPyuOPP26Wp86ePWs6C+fLl887ZwjAI7QJn1byaQPMHj16mDw5ALCiHNcnawKiBjUAApfOrjqbVWpp97333isxMTEsQwGwNLeDGy271tmbq1m48M/GdQD8S9smfPfdd5KQkOB6I5Ld5pcAYKvgRquZMnYi1B4zv/32mynpBuBf+jepFYfOvlO//PILs6wAbMXt4Obtt9/O9Prhw4eb/BsA/t2vbdKkSXLkyBFzrAn/LVu25FcCwFY8tifA3XffbZrwvfnmm566SwBu0BmaWbNmmZmbqKgos5N3TnpPAUCw81hwo71utIMwAN87ePCgTJ8+3XyuG9hqYEMFIwC7cju40SfNtBwOh3liXb16dY6b+AG4PrGxsaYhn/acatGiRVB39QYAnwc3GXtj6JNotWrV5MUXX5T27dtf9wkBuDZ9U6HLULrsFB0dba7j7w8AchDc6J40uqFlnTp1TK8MAL6nHYY1t0ZLvcuWLWuqFJmpAYAcBje6/4y+O9y0aRPBTRYcDndGFcg+3VxW94b6448/TL+pKlWqZNl3CgDsyO1lqdq1a8uOHTtM0iIyD2zuGLuKoYHHl6HWrFlj+tfoDKouRekWCjpzAwC4zuDm5ZdfNptkvvTSS9KgQQNTcpqWc/3fri5eFtl06Iz5vGZstESEsdsyrn8ZaubMmbJhwwZzrJtdatdh57YKAIAcBjeaMPyPf/xDOnToYI67dOmSbjpc31nqsb6rxJ8mDWzCkgGum+bTHD161Hxs06aNqYpiKQoAPBDcjBgxQgYOHCiLFi3K7pfYHqkQyCl9s6A0iAkLC5OePXuaGZzSpUszqADgqeDG+WRLK3fAuy5cuCAzZswwvWu0Z40qWrQoww4A3si5YSoc8K79+/ebaqiTJ0/Ktm3b5IYbbqDTMAB4M7jRRMZrBTjHjx939xwA29OZ0RUrVsj3338vly9fNq0WdCmKLRQAwMvBjebdZOxQDOD6nD9/3uwLtXXrVnNcs2ZN6dy5M3u1AYAvgpvbb79dihUrltPvBSADrS4cO3asmfHUJpnx8fHSsGFDloABwBfBDfk2gOdpQPO3v/3NLEn16tVLSpQowTADgK+rpQBcn6SkJDl37pyrAkpnauLi4kzJNwDAh8GNJjkCuD67d++WKVOmSGhoqAwYMMDk1Th72QAA/LT9AgD36cznkiVLZPHixebzIkWKmBkcDW4AAJ5FcAN42dmzZ2XatGlmw1lVr149s41JeHg4Yw8AXkBwA3jRzp07ZerUqSbA0aUnDWo0vwYA4D0EN4AXaRWUBjaaPKzVUGyjAADeR3ADeFFCQoIsXbpUWrVqRdIwAPhIbl99I8AOtm/fLomJia7jyMhIad++PYENAPgQMzeAB2irhEWLFplZGlWmTBmpUaMGYwsAfkBwA1yn06dPm941e/bsMccNGjSQypUrM64A4CcEN8B12LZtmynz1s0vtbS7S5cuUqtWLcYUAPyI4AbIIW3Kt3DhQvN5bGys9OzZUwoVKsR4AoCfEdwAOaQBjWrUqJG0a9fObKkAAPA/no0BN+iGl1FRUeZzzasZNGgQvWsAIMBQCg5kQ2pqqsydO1fee+89OXHihOt6mvIBQOAhuAGuQYOZcePGyc8//ywXLlwwScQAgMDFshSQhY0bN8qMGTMkOTlZIiIiTMfhatWqMWYAEMAIboBMXLp0yXQaXrVqlaspX48ePaRAgQKMFwAEOIIbIBO6BOUMbJo1a2b2hgoJCWGsACAIENwAmWjcuLHs2rXLlHlXqVKFMQKAIEJCMSAiKSkpsmzZMrNHlNKeNXfddReBDQAEIWZuYHvHjh2TSZMmyZEjR0w1VOvWrW0/JgAQzAhuYGu//PKLzJo1y8zcaHO+8uXL+/uUAADXieAGtnTx4kWZM2eOrF+/3hxXqFBBunfvLvny5fP3qQEArhPBDWzn6NGjZhlKP+bKlUtatmwpLVq0kNy5SUEDACsguIHtOBwO03VYZ2m0dw1LUQBgLQQ3sAWtgnLOzBQrVkx69+5tdvV2boIJALAO5uFheYcOHZIxY8bInj17XNfpjt4ENgBgTQQ3sPTy0+rVq2Xs2LEmv2b+/PnmOgCAtbEsBUvSjS5nzpwpGzZsMMfaZbhr164mgRgAYG0EN7CcgwcPyuTJk+X48eMmz6ZNmzbSpEkTAhsAsAmCG1iKdhn+5JNPJDU11ezgrdVQuqM3AMA+CG5gKUWLFpWqVaua6qiEhASJiIjw9ykBAHyM4AZB78CBA1KoUCHJmzevWXrq1q2b2fiS/BoAsKeAqJYaPXq0aaSmL06NGzeWlStXXvW2H3/8sekmGxMTYy5t27bN8vawLq18Wr58uVmG+u6771yVUGFhYQQ2AGBjfg9uvv76axk6dKgMGzZM1q5dK/Xq1ZP4+HiTO5GZxYsXyx133CGLFi0yL2yaT9G+fXvZv3+/z88d/nPp0iWTNJyYmGiWoDSw0TwbAAD8HtyMGjVK+vfvL/fdd5/UrFnTNFuLjIyUcePGZXr7L7/8UgYNGiRxcXFSvXp108NEX9wWLFjg83OHf+zbt0+2bNki27Ztk5CQEOnQoYP07NnTLEUBABDq752Z16xZI08//bTrOi3d1aUmnZXJjqSkJElJSTE5F1frd6IXp9OnT5uP+jV68fTP8+5vIa5j8z1y0TTOU3R2ZsWKFWb2Tj8vWLCg2cm7RIkSZiYHnuX8+/D03wkYZ3/hMR3c4+zO/fk1uDl27JhZSihevHi66/V48+bN2bqPJ598UkqWLGkCosyMHDlSRowYccX1upyhM0SelJwqsj/pzyEtFemQRfMThZ5xnqMBjM7YOAMbXZLUpUx4l3Z2hvcxzr7DWAfnOOtkRnYF9Tz+a6+9JhMnTjTv5DUZOTM6K6Q5PWlnbpx5OtHR0R49n1PnLois/NF8PmtoG4nKE9TDG5B0fyjNxzp8+LD5HWryMLxD3yXpk1O7du0YZy9inH2HsQ7ucXauvGSHX199ixQpYnIm9IUqLT3WpYasvPnmmya4+f7776Vu3bpXvV2ePHnMJSMdcE+/MIaF/bU0Eh6u909wcz10hmbJkiVmlsb5O65UqZKULVtWZs+e7ZXfIa7EOPsG4+w7jHVwjrM79+XXhOLw8HBp0KBBumRgZ3Kwtsu/mtdff11eeuklmTt3rjRs2NBHZwtfOnv2rHzxxRemKk7LvN2J2AEA9ub3qQVdMurbt68JUho1aiTvvPOOnDt3zlRPqT59+kipUqVM7oz617/+JS+88IJ89dVXpjfOoUOHzPX58uUzFwS/nTt3ytSpU02AoxVQt956q+TPn9/fpwUACBJ+D2569+4tR48eNQGLBipa4q0zMs4kY82x0Aoqpw8++MBUJWnpb1raJ2f48OE+P394js7a/fjjj+aiS1K6lUKvXr3MRwAAgia4UYMHDzaXzGiycFq7du3y0VnB14GNLkPprI264YYbzIwNOTUAgKAMbgCdndOSfm3Q16lTpyyTxAEAyArBjQf9b2sjuDFbc/78eYmKijLHrVq1kvr161+1ISMAAEGx/YJVaI7IHWNX+fs0goZWP3322WcmMdy5J5S2BSCwAQBcL2ZuPOR8SqpsOnTGfF6jRH6JCPtrGwakp3tCTZs2zczaaDsAbcoXGxvLMAEAPILgxgv+0+9GycW+C1fQGZqFCxfKsmXLzLEGNFr1xmwNAMCTCG68gLjmSidPnpQpU6aYhGGlPY20NTc7eQMAPI3gBj4xc+ZME9joVhgJCQlSo0YNRh4A4BUEN/CJjh07yqxZs0yZd0xMDKMOAPAaqqXgFSdOnJC1a9e6jjWv5p577iGwAQB4HTM38LiNGzfKjBkzJDk52ezoXbFiRUYZAOAzBDfwmEuXLkliYqKsWvVnv5/SpUtTCQUA8DmCG3jE8ePHZdKkSa5d2ps2bSqtW7c2jfkAAPAlghtctw0bNphlKN2tPSIiQrp16yZVqlRhZAEAfkFwg+umQY1eypYtKz169JDo6GhGFQDgNwQ3yPGml7qTt4qLizPbKGjvGud1AAD4C69EcNsvv/wiH3zwgSQlJZlj3WqiVq1aBDYAgIBAcINs06Wnb7/9VqZPny7Hjh2Tn3/+mdEDAAQclqWQLbpz9+TJk+Xo0aPmuGXLlnLTTTcxegCAgENwgyw5HA5Zv369zJ492/SxyZcvn3Tv3l0qVKjAyAEAAhLBDbKkDfnmzJljPtdOw1rmrQEOAACBiuAGWapbt67JrdGKqObNm5vkYQAAAhnBDa5YhtqxY4eZpdFAJm/evPLQQw9JaCgPFQBAcKBaCi660eXUqVPliy++SLejN4ENACCY8HYcxsGDB001lO4RpY34UlJSGBkAQFAiuLE5XYbSpGHdzTs1NVUKFChgtlAoU6aMv08NAIAcIbixsQsXLpgNLzdt2mSOq1WrJgkJCWbzSwAAghXBjY0dPnxYNm/ebJah2rVrJ40bN6YaCgAQ9AhubKxcuXJy6623SsmSJaVUqVL+Ph0AADyCaikbOX/+vEyZMsXsC+V04403EtgAACyFmRub2Lt3rwlsTp06ZSqi+vXrxxIUAMCSCG5sUA21bNkyWbhwoVy+fFliYmKkU6dOBDYAAMsiuLGwpKQkmT59umzbts0c16pVSzp37ix58uTx96kBAOA1BDcWpUtP48ePlzNnzpgOw7fccovUr1+fGRsAgOUR3FiUNuMrWLCghIeHS69evaR48eL+PiUAAHyC4MZCzp07Zza6DAkJMRcNanQJSgMcAADsglJwi9i5c6eMGTNGFixY4Louf/78BDYAANshuAlyWgG1ePFimTBhgpw9e1Z+//13Nr0EANgay1JBTJOFp02bZmZtVFxcnHTo0EHCwsL8fWoAAPgNwU2Q2r59uwlsNM9Gg5mOHTtKvXr1/H1aAAD4HcFNkO7mPWnSJElOTpZixYqZxOEiRYr4+7QAAAgIBDdBSCuitMuwLkdp/xqWoQAA+AvBTZDQLsPajK9ChQrmuHbt2uYCAADSI7gJcKmpqWZfKN0fKioqSgYOHCj58uXz92kBABCwCG4CmO7gPXnyZNm3b585rlmzplmSAgAAV0dwE6C2bNliNr3U5GHtMtylSxcT3AAAgKwR3ARgU7758+fLihUrzHHJkiWlZ8+eEhMT4+9TAwAgKBDcBJhcuXKZ3jWqcePG0q5dO7NPFAAAyB6CmwCascmdO7cJbrQhX506daRKlSr+Pi0AAIIOe0v52aVLl2T27NnyzTffiMPhMNdpjg2BDQAAOcPMjR8dP37cVEMdPHjQHO/Zs0fKlSvnz1MCACDoEdz4yW+//SYzZ86UixcvSkREhHTt2pXABgAADyC48bGUlBSZN2+erFmzxhyXLVtWevToIdHR0b4+FQAALIngxsemTJlietio5s2bS6tWrUwiMQAA8AyCGx/TgObAgQOSkJAglSpV8vW3BwDA8ghufLAMtX//filfvrw5Ll26tDz66KNmE0wAAOB5rId40dGjR+Xjjz+WL7/8Ug4fPuy6nsAGAACLBzejR482Mxu6KaR25V25cmWWt580aZJUr17d3F6b3WmfmECi/WrWrVsnH330kQlw9DyTk5P9fVoAANiC34Obr7/+WoYOHSrDhg2TtWvXSr169SQ+Pl6OHDmS6e2XLVsmd9xxhzzwwAMmgNASar1oaXUgCJVUSZwzW2bMmGEa9FWsWFEefPBBUxUFAABsENyMGjVK+vfvL/fdd5/Z9XrMmDESGRkp48aNy/T27777rtxyyy3y+OOPS40aNeSll16S+vXry3vvvSf+FpMrSTrn2SSbNm4w2yhoJdTdd98t+fLl8/epAQBgG37NatUGdtrv5emnn3Zdp2XRbdu2leXLl2f6NXq9zvSkpTM906dPz/T2uhyUdkno9OnTrkRfvXhKSsolKRtyUgrmviBRUVHSrVs3M1ujszfwLOfvzZO/PzDO/sLjmbG2mhQvPUe7c39+DW6OHTsmqampUrx48XTX6/HmzZsz/ZpDhw5lenu9PjMjR46UESNGXHF9YmKimSHylORUkf9eipXc4pBeZQubZbJAWSqzqvnz5/v7FGyBcWacrYbHdHCOc1JSUrZva/l6ZJ0VSjvTozM3ZcqUkfbt23u0K7AmEbdunSwLFy6UjvFtJTw83GP3jSujd/2jadeunYSFhTE8XsI4+wbj7DuMdXCPs3PlJeCDmyJFikhISEi6MmmlxyVKlMj0a/R6d26vO2zrJSMdcE+/MBbIlUvyhIgJbHjR9T5v/A7BOPsLj2fG2mrCPPwc7c59+TWhWIOABg0ayIIFC1zXXb582Rw3adIk06/R69PeXmmEeLXbAwAAe/H7spQuGfXt21caNmwojRo1knfeeUfOnTtnqqdUnz59pFSpUiZ3Rj322GPSsmVLeeutt6Rjx44yceJEWb16tekpAwAA4Pfgpnfv3qbR3QsvvGCSguPi4mTu3LmupOE9e/ak21iyadOm8tVXX8lzzz0nzzzzjFSpUsVUStWuXduPPwUAAAgUfg9u1ODBg80lM4sXL77iul69epkLAABAwDXxAwAA8CSCGwAAYCkENwAAwFIIbgAAgKUQ3AAAAEshuAEAAJZCcAMAACyF4AYAAFgKwQ0AALCUgOhQ7EsOh8PtrdPd2eY9KSnJ3De7VXsP4+wbjDPjbDU8poN7nJ2v287X8azYLrg5c+aM+VimTBl/nwoAAMjB63iBAgWyvE0uR3ZCIAu5fPmyHDhwQPLnzy+5cuXy6H1rVKlB0969eyU6Otqj9w3G2dd4PDPOVsNjOrjHWcMVDWxKliyZbkPtzNhu5kYHpHTp0l79HvrLJLjxPsbZNxhnxtlqeEwH7zhfa8bGiYRiAABgKQQ3AADAUghuPChPnjwybNgw8xHewzj7BuPMOFsNj2n7jLPtEooBAIC1MXMDAAAsheAGAABYCsENAACwFIIbAABgKQQ3bho9erSUL19e8ubNK40bN5aVK1dmeftJkyZJ9erVze3r1Kkjs2fPvp7fl224M84ff/yxtGjRQmJiYsylbdu21/y9wP1xTmvixImmw3fXrl0ZSg8/ntXJkyfl4YcfltjYWFNxUrVqVZ47vDDO77zzjlSrVk0iIiJMR90hQ4bIhQsXeExn4ccff5TOnTubLsH6HDB9+nS5lsWLF0v9+vXNY7ly5coyfvx48TqtlkL2TJw40REeHu4YN26cY8OGDY7+/fs7ChYs6Dh8+HCmt//pp58cISEhjtdff92xceNGx3PPPecICwtz/Prrrwy5B8f5zjvvdIwePdqxbt06x6ZNmxz33nuvo0CBAo59+/Yxzh4cZ6edO3c6SpUq5WjRooUjISGBMfbwOCcnJzsaNmzo6NChg2Pp0qVmvBcvXuxYv349Y+3Bcf7yyy8defLkMR91jOfNm+eIjY11DBkyhHHOwuzZsx3PPvusY+rUqVpp7Zg2bVpWN3fs2LHDERkZ6Rg6dKh5Hfz3v/9tXhfnzp3r8CaCGzc0atTI8fDDD7uOU1NTHSVLlnSMHDky09vfdtttjo4dO6a7rnHjxo4HH3wwp78vW3B3nDO6dOmSI3/+/I7PPvvMi2dpz3HWsW3atKlj7Nixjr59+xLceGGcP/jgA0fFihUdFy9edO8XanPujrPetnXr1umu0xfgZs2aef1crUKyEdw88cQTjlq1aqW7rnfv3o74+HivnhvLUtl08eJFWbNmjVnySLtPlR4vX74806/R69PeXsXHx1/19sjZOGeUlJQkKSkpUqhQIYbUg49n9eKLL0qxYsXkgQceYGy9NM4zZsyQJk2amGWp4sWLS+3ateXVV1+V1NRUxtyD49y0aVPzNc6lqx07dpilvw4dOjDOHuSv10HbbZyZU8eOHTNPLvpkk5Yeb968OdOvOXToUKa31+vhuXHO6MknnzTrwRn/oHB947x06VL55JNPZP369QylF8dZX2QXLlwod911l3mx/f3332XQoEEmYNeur/DMON95553m65o3b252m7506ZIMHDhQnnnmGYbYg672Oqg7h58/f97kO3kDMzewlNdee80ku06bNs0kFcIzzpw5I/fcc49J3i5SpAjD6kWXL182s2MfffSRNGjQQHr37i3PPvusjBkzhnH3IE1y1Rmx999/X9auXStTp06VWbNmyUsvvcQ4WwAzN9mkT+ghISFy+PDhdNfrcYkSJTL9Gr3endsjZ+Ps9Oabb5rg5vvvv5e6desynB58PG/fvl127dplqiTSvgir0NBQ2bJli1SqVIkxv85xVlohFRYWZr7OqUaNGuYdsC6/hIeHM84eGOfnn3/eBOz9+vUzx1rNeu7cORkwYIAJJnVZC9fvaq+D0dHRXpu1Ufz2skmfUPRd1IIFC9I9ueuxro9nRq9Pe3s1f/78q94eORtn9frrr5t3XHPnzpWGDRsylB5+PGs7g19//dUsSTkvXbp0kVatWpnPtYwW1z/OqlmzZmYpyhk8qq1bt5qgh8DGM49nZ25exgDGGVCy5aLn+O110KvpyhYsNdTSwfHjx5uStgEDBphSw0OHDpn/v+eeexxPPfVUulLw0NBQx5tvvmlKlIcNG0YpuBfG+bXXXjMloJMnT3YcPHjQdTlz5oznHwQ2HueMqJbyzjjv2bPHVPsNHjzYsWXLFsd3333nKFasmOPll1++zt+4tbk7zvp8rOP8n//8x5QrJyYmOipVqmSqXHF1+ryqbTf0oiHEqFGjzOe7d+82/69jrGOdsRT88ccfN6+D2raDUvAApDX6ZcuWNS+mWnq4YsUK1/+1bNnSPOGn9c033ziqVq1qbq/lcLNmzfLDWVt7nMuVK2f+yDJe9MkLnhvnjAhuvPN4VsuWLTNtI/TFWsvCX3nlFVOGD8+Nc0pKimP48OEmoMmbN6+jTJkyjkGDBjlOnDjBMGdh0aJFmT7fOsdWP+pYZ/yauLg483vRx/Onn37q8LZc+o9354YAAAB8h5wbAABgKQQ3AADAUghuAACApRDcAAAASyG4AQAAlkJwAwAALIXgBgAAWArBDQAAsBSCGwDpjB8/XgoWLBi0o5IrVy6ZPn16lre59957pWvXrj47JwC+RXADWJC+eOuLfMaLbsgYCMGT83x048LSpUvLfffdJ0eOHPHI/R88eFBuvfVW87nuZK7fRzf3TOvdd9815+FNw4cPd/2cuiGjbi6qO04fP37crfshEAPcF5qDrwEQBG655Rb59NNP011XtGhRCQTR0dGyZcsWs3PzL7/8YoKbAwcOyLx58677vkuUKHHN2xQoUEB8oVatWvL9999LamqqbNq0Se6//345deqUfP311z75/oBdMXMDWFSePHnMC33ai84gjBo1SurUqSNRUVFmNmHQoEFy9uzZq96PBh+tWrWS/Pnzm6CkQYMGsnr1atf/L126VFq0aCERERHm/h599FE5d+5cluemsxl6PiVLljSzLPo1GgScP3/eBDwvvviimdHRnyEuLk7mzp3r+tqLFy/K4MGDJTY2VvLmzSvlypWTkSNHZrosVaFCBfPxhhtuMNfffPPNV8yGfPTRR+Y89PumlZCQYIIRp2+//Vbq169vvmfFihVlxIgRcunSpSx/ztDQUPNzlipVStq2bSu9evWS+fPnu/5fg54HHnjAnKeOX7Vq1cysUtrZn88++8x8b+cs0OLFi83/7d27V2677TazhFioUCFzvjpTBYDgBrAdXQr6v//7P9mwYYN54Vy4cKE88cQTV739XXfdZQKNVatWyZo1a+Spp56SsLAw83/bt283M0Q9evSQ//73v2ZGQoMdDT7coS/sGlxosKAv7m+99Za8+eab5j7j4+OlS5cusm3bNnNbPfcZM2bIN998Y2Z/vvzySylfvnym97ty5UrzUQMnXa6aOnXqFbfRgOOPP/6QRYsWua7TpSMNqPRnV0uWLJE+ffrIY489Jhs3bpQPP/zQLGu98sor2f4ZNfDQmanw8HDXdfoz69hOmjTJ3O8LL7wgzzzzjPnZ1D//+U8TwOgY6/nrpWnTppKSkmLGRQNOPbeffvpJ8uXLZ26nwR9ge17fdxyAz/Xt29cREhLiiIqKcl169uyZ6W0nTZrkKFy4sOv4008/dRQoUMB1nD9/fsf48eMz/doHHnjAMWDAgHTXLVmyxJE7d27H+fPnM/2ajPe/detWR9WqVR0NGzY0xyVLlnS88sor6b7mxhtvdAwaNMh8/sgjjzhat27tuHz5cqb3LyKOadOmmc937txpjtetW3fF+CQkJLiO9fP777/fdfzhhx+a80hNTTXHbdq0cbz66qvp7mPChAmO2NhYx9UMGzbMjIOOfd68ec156GXUqFGOrDz88MOOHj16XPVcnd+7WrVq6cYgOTnZERER4Zg3b16W9w/YATk3gEXpUtIHH3zgOtZlKOcshi7jbN68WU6fPm1mSy5cuCBJSUkSGRl5xf0MHTpU+vXrJxMmTHAtrVSqVMm1ZKWzKzp74qTxhc5I7Ny5U2rUqJHpuWneic406O30ezdv3lzGjh1rzkdzb5o1a5bu9nqs38u5pNSuXTuzhKMzFZ06dZL27dtf11jpDE3//v3l/fffN0th+vPcfvvtZpbL+XPq7EjamRpdUspq3JSeo84y6e2++OILk9j8yCOPpLvN6NGjZdy4cbJnzx6zLKczL7oUlxU9H00O15mbtPT76GwaYHcEN4BFaTBTuXLlK5ZGNBh46KGHzAu15mroMpLmfeiLamYv0pr3ceedd8qsWbNkzpw5MmzYMJk4caJ069bN5Oo8+OCDJmcmo7Jly1713PRFee3atSZ40NwZXZZSGtxci+a9aOCk56KBmi7baNA1efJkyanOnTuboEx/xhtvvNEs9bz99tuu/9efU3NsunfvfsXXag7O1egSlPN38Nprr0nHjh3N/bz00kvmOh1HXXrSZbgmTZqYcXnjjTfk559/zvJ89Xw09yltUBloSeOAPxHcADaiOTM6W6Ivps5ZCWd+R1aqVq1qLkOGDJE77rjDVGFpcKOBhuaKZAyirkW/d2ZfownLmtyrsyQtW7Z0Xa/HjRo1Sne73r17m0vPnj3NDI7myWiwlpYzv0VnWbKiAYoGLhos6IyIzrjoz+akn2t+j7s/Z0bPPfectG7d2gSXzp9Tc2g0qdsp48yL/gwZz1/PR/ObihUrZsYCQHpUSwE2oi/Omoz673//W3bs2GGWmsaMGXPV2+syiSYHa4XO7t27zYuxJhY7l5uefPJJWbZsmbmNLrlo0q9W9ribUJzW448/Lv/617/Mi7cGFJrArPetybxKq73+85//mGW1rVu3mmRcrUjKrPGgvvjrrJAmBx8+fNgsh2W1NKUzN7pE5EwkdtJE388//9zMumgitpZ166yLBivu0NmZunXryquvvmqOq1SpYirPNNFYf5bnn3/ejG9amiytS386FseOHTO/Pz2/IkWKmAopnWXSmSz9HekM2r59+9w6J8CS/J30A8DzMktCddKEVk2E1eTT+Ph4x+eff24SXU+cOHFFwq8mqd5+++2OMmXKOMLDw02S7eDBg9MlC69cudLRrl07R758+UzybN26da9ICM4qoTgjTeIdPny4o1SpUo6wsDBHvXr1HHPmzHH9/0cffeSIi4sz3ys6Otok+65duzbThGL18ccfm/PX5N6WLVtedXz0++q46Ndv3779ivOaO3euo2nTpmbc9Ps2atTInEtWCcV67hn95z//ceTJk8exZ88ex4ULFxz33nuvGY+CBQs6HnroIcdTTz2V7uuOHDniGl89t0WLFpnrDx486OjTp4+jSJEi5v4qVqzo6N+/v+PUqVNXPSfALnLpP/4OsAAAADyFZSkAAGApBDcAAMBSCG4AAIClENwAAABLIbgBAACWQnADAAAsheAGAABYCsENAACwFIIbAABgKQQ3AADAUghuAACAWMn/A4JeivoNMVHGAAAAAElFTkSuQmCC",
"text/plain": [
"<Figure size 640x480 with 1 Axes>"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"# ROC AUC\n",
"from sklearn.metrics import roc_curve\n",
"import matplotlib.pyplot as plt\n",
"\n",
"fpr, tpr, thresholds = roc_curve(df_test[\"target\"], y_pred_proba[:, 1])\n",
"plt.plot(fpr, tpr)\n",
"plt.plot([0, 1], [0, 1], linestyle='--', color='gray')\n",
"plt.xlabel(\"False Positive Rate\")\n",
"plt.ylabel(\"True Positive Rate\")\n",
"plt.grid()"
]
},
{
"cell_type": "code",
"execution_count": 225,
"id": "9d0f7bd1",
"metadata": {},
"outputs": [
{
"data": {
"image/png": "iVBORw0KGgoAAAANSUhEUgAAAiMAAAGdCAYAAADAAnMpAAAAOnRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjEwLjgsIGh0dHBzOi8vbWF0cGxvdGxpYi5vcmcvwVt1zgAAAAlwSFlzAAAPYQAAD2EBqD+naQAAXsxJREFUeJzt3Qd4lGW6N/B/ek8gCSmQEELohN57kSIgSBUVQV0FbGc9+p09u66uLOvZdc+66+rZRbE3VAQEpReRKk0ILXQIJJCekN6Tme+6n2RiEpIwSabP/3ddQ6a8886bN0Pmzv3cz/04aLVaLYiIiIjMxNFcL0xEREQkGIwQERGRWTEYISIiIrNiMEJERERmxWCEiIiIzIrBCBEREZkVgxEiIiIyKwYjREREZFbOsAIajQZJSUnw8fGBg4ODuQ+HiIiI9CB9VfPy8tC2bVs4OjpadzAigUh4eLi5D4OIiIia4ebNmwgLC7PuYEQyIrpvxtfX12D7LSsrw86dOzFp0iS4uLgYbL/E82wufE/zPNsSvp+t/zzn5uaqZILuc9yqgxHd0IwEIoYORjw9PdU+GYwYD8+z6fBc8zzbEr6fbec8363EggWsREREZFYMRoiIiMisGIwQERGRWTEYISIiIrNiMEJERERmxWCEiIiIzIrBCBEREZkVgxEiIiIyK6toekZERLalQqPFseu3kZZXjCAfdwyO9IeTo4NFv05L9mWq79dugpH9+/fjjTfewIkTJ5CcnIwNGzZg5syZjT5n7969ePHFF3Hu3DnVFvaVV17BY4891pLjJiIiK7U9NhnLN51Hck5x9X2hfu5YNr0H7o0OtcjXacm+TPX92tUwTUFBAfr06YMVK1botf3169cxbdo0jBs3DqdOncJ//ud/4sknn8SOHTuac7xERGTF5IP56VUxtT6YRUpOsbpfHre012nJvkz1/dpdZmTKlCnqoq+VK1ciMjIS//jHP9Tt7t274+DBg/jnP/+JyZMnN/XliYjISslQhWQItPU8JvfJoIU8PrFHSIuGMAz5Oi3Zl6m+X1tg9JqRw4cPY8KECbXukyBEMiQNKSkpUZeaq/7pFvORi6Ho9mXIfRLPsznxPc3zbMmOXr99R4ag7ge0PP7MquMI8XOHRqNBQrwjjm8+D0dH/RP5knVoyusYa1/6Pvfw1TQMifSHLf7e0HefRg9GUlJSEBwcXOs+uS0BRlFRETw8PO54zuuvv47ly5ffcb8scSwrCxrarl27DL5P4nk2J76neZ4t0YkM+evf6a7b7TifVuOWI5ByyyjHU/t1zLevnQeOIvNCffkT6/+9UVhYaL2zaV566SVV8KojgYsUvk6aNEktcWzIiE1O/sSJE422bDLxPJsS39M8z5Ys4PptfH7l+F23m9E7BO1aeaBCo1F1hzLU79SEzEhidhE2nknR+3WMtS99nztp1BCTZ0Zu376NY8eOqc/ViooKo30W6kY2zB6MhISEIDU1tdZ9cluCivqyIsLNzU1d6pKTZIygwVj7JZ5nc+F7mufZEg3rFKRmkcjwRX15AMmbyFDHPx/sr2ooJLjeuvUapk7u2qTf0VKr8XP8j3q/jrH2pe9zh3UKMmnNSGxsLDZt2oTS0lL1WTx8+HCj/d7Qd39Gb3o2bNgw7N69u9Z9EoHJ/UREZD/kA1ems9ZH91Esj7f0g7nm6zi08HVasi9DHochSHC3efNmfPvttyoQad++Pfr16wdL0ORgJD8/X03RlYuQFJpcT0hIqB5iWbRoUfX2Tz31FOLi4vDf//3fuHjxIt555x2sWbMGL7zwgiG/DyIisgLSV+P/Hup7x/2SIXj3kf4G67sh+5H91S0qbc7rtGRfhjyOlsjIyMBHH32keoSJUaNG4dFHHzVo6UNLNHmY5vjx46pniI6utkO+qU8//VQ1QtMFJkLG+rZs2aKCj7fffhthYWH48MMPOa2XiMhOhfpVDtH7ujvjtfujEeRrnI6k8kEv02YN0fm0Jfsy5HE0x6VLl1Q2RDIjXl5emDVrFqKiomBJmhyMjB07Flptw1W/EpDU95yTJ082/eiIiMjmHLtxW30dHhWI+/u1M+pryQf+sKgAs+/LkMfRVP7+lcWxHTp0wOzZs+Hj4wNLY5GzaYiIyHZJhkAMMmNvDVtXXFwMd/fKoaE2bdrg8ccfV201mtKvxZQs86iIiMgmyQyTEzey1HVzNvqyVVqtVo1ESJfzmiUToaGhFhuICGZGiIjIZC4k5yKvpBzebs7oHmoZxZO2orS0VNVonjlzRt2WoERmzFgDBiNERGQyP1fVi/SPaG3367EYUmpqKtauXYvMzEw4ODioiSYjR46EtWAwQkREJg9GOERjuGGZmJgYbN++HeXl5ao4dc6cOYiIiIA1YTBCREQm++CsLl7twHoRQ7h27ZpqZCY6d+6MmTNnGmUNN2NjMEJERCZxPaMAGfmlcHV2RO8wP551A4iKikJ0dLRaekXaussQjTViMEJERCYdoukb1gruLndfvZfqzy5J1/Pu3burqbsSfEjvEGsNQnQsd54PERHZlKPV/UVam/tQrLZ3yLp167Bx40a10J2uAam1ByKCmREiIjJpZmRwpHk6kVqzxMREFYhkZ2erfiHh4eGwJQxGiIjI6JJzinDzdhFkOZb+7VvxjOtJq9Xi6NGjarV7jUaDVq1aYe7cuWjXzrht9E2NwQgRERmdbhZNz7Z+8HF34RnXQ1FREb7//nu10J2QOpEZM2ZUt3m3JQxGiIjIZEM0nNKrP41Gg6SkJDg5OWHSpEkYNGiQTdSH1IfBCBERmSwzMpjFq3cdlnGoCji8vLwwb948ODs7q7VlbBln0xARkVFlFZTicmq+us7MSMMKCwvx9ddf4/Tp09X3SaGqrQcigpkRIiIyquPxlav0RrXxQoC3G892PeLj4/Htt98iLy8Pt27dUvUhrq6udnOuGIwQEZFRHbueqb4OjmQL+PqGZQ4ePIg9e/ao6wEBAWpoxp4CEcFghIiIjOrYjcrMCIOR2goKCrBhwwa1vozo3bs3pk2bZneBiGAwQkRERlNQUo7YxBx1nfUivygpKcF7772nhmWcnZ0xdepU9O3b12Zny9wNgxEiIiOr0FSuVpuWV4wgH3eVIXCS7l924GRCtvr+27XyQFhr61tN1ljc3NzQp08f1UNk3rx5aNOmDewZgxEiIiPaHpuM5ZvOIzmnuPq+UD93LJveA/dG2/4siWPV/UW4Hk1+fj7Ky8tVF1Uxbtw4jB49Gi4ubALHqb1EREYMRJ5eFVMrEBEpOcXqfnncXopXB9l58WpcXBxWrlyJtWvXoqKiQt0na8wwEKnEYISIyAhkaEIyIpXrqtamu08el+1sVWm5Rg3TiCF2GoxIF9Uff/wRX3zxhSpYlUBEvlJtHKYhIjICqRGpmxGpSUIQeVy2GxZlm6vYnk3MRkm5Bv5erohq4w17k5ubi/Xr16seImLAgAGYPHkysyH1YDBCRGQEUqxqyO2s0bHrWdX1IvY2S+TKlSv47rvvVFdVmao7ffp0REdHm/uwLBaDESIiI5BZM4bczhrZ6+J40rxs7969KhAJCQlRs2X8/e3rHDQVgxEiIiOQ6bsya0aKVeurCpE8QYhf5TRfWyS1MLpgxFa/x4ZIFmjOnDk4fvw4xo8fr/qIUONYwEpEZATSR0Sm7zZGHrfVfiOXUvKQV1wOL1cn9Aj1ha2TfiE//fRT9W3JhEyaNImBiJ4YjBARGYn0EXn3kf5wrifg+POsaJvuM6LLivSPaA1nJ9v9qJHZMTt27MDq1avxww8/ICEhwdyHZJVs9x1CRGQBhkQGoLxq+u7/zIxG91AfdT27qAy2TGYJicE2XC+SlZWFTz75BEeOHFG3hwwZgnbt2pn7sKwSgxEiIhN0IO0U5I1Hhkbg8eGR6vb6mERV6GiL5Ps6ZuP1IhcuXFBryyQmJsLd3R0PPvgg7r33Xjg5OZn70KwSgxEiIiM6Gne7VtOvKb1C4ObsiKtp+ThbtYCcrbmRWYj0vBK4OjmiT3hl63NbsmvXLqxZs0YtdhcWFoalS5eia9eu5j4sq8ZghIjIiI5WtUMf0rGysZmPuwsm9wypzo7Yop+rhmj6hPvB3cX2MgWBgYHq6/Dhw/HYY49VrzVDzcdghIjISHIKy3A+OVddH1pjuGJW/8q6go2nk1BWobHhxfFsZ4imqKio+nrfvn1VNmTixIkcljEQBiNEREacUSJlIZGBXgjy/aW52ahOgQj0dsPtglLsu5Rus8WrtrA4XllZGTZv3qwWuZMmZro+ItLMjAyHwQgRkbGHaOp8KMtU1/v7tlXXN5y0raGa1NxiJNwuhMxmHhDRGtYsIyMDH330EU6cOKHWmbl69aq5D8lmMRghIjKSo1UZgiEd78wQzK4aqtl1IVUN59haVqR7qC983V1grc6cOYP3338fqamp8PT0xCOPPILevXub+7BsFnvUEhEZQV5xGWKrZstIr5G6pCtptxAfXEzJw5azyXh4SHvb6i9ipUM0Miyzbds2nDx5Ut3u0KEDZs+eDR+fyv4wZBzMjBARGcHx+CxIr7Nwfw+0beVxx+NSdzCrX2V2ZMPJWzbzM6hej8ZKi1f37NlTHYiMGTMGCxcuZCBiAgxGiIiM2F9kaD1ZEZ2Z/dqp2oqfb2QhPrPA6n8O2YWlKtMjBlppMDJ69GiEh4dj0aJFGDt2LBwd+TFpCjzLREQm6C9Sn2Bfd4zoFGgzhazHb2Sprx3beKGNjxusQWlpKX7++efqbrjSTfXxxx9HZGRlp1wyDQYjREQGVlhajrO3dPUijWcIdIWsEoxYe3t4axuikeLUDz74AFu3blUzZmoOoZFpMRghIjKwE/FZanG8dq08EO7v2ei20o3V09UJ8ZmFiEmozCxY++whS292JkGfBB8ffvihmr4rxalt2rQx92HZNQYjRERGXo+mMZ6uzpgSHaquf2vF7eElG6SbPWTJM2lkPZn169erRmbl5eXo1KkTnnrqKURERJj70OwagxEiIqPVi+j3oawbqtl8Ogkl5RVW+fM4lZCtskGhfu4Ia33n7CFLkJKSonqHxMbGqqGYCRMm4OGHH1Z9RMi8GIwQERlQUWkFTt9suL9IfYZ2DFAf4rnF5fjxQppVD9FIVsRSay6Ki4uRlZUFX19fVaQ6YsQIiz1We8NghIjIgE4mZKG0QoNgXzdEBOj3F7eTo4Oa5mvNQzW64lVLqxepWRQsDczmzJmjhmVk+i5ZDgYjREQGdETXAj4yoEl/dc+uCkb2XkpTC+hZk9JyTXXxrSXViyQlJeG9995TRao6PXv2hIeHZQ4j2TMGI0REBnQ0rmn1Ijqdg33Qq52fqrvYdDrJqn4msUk5KC7ToLWnCzq18baIbMiRI0fUIncyffeHH34w9yHRXTAYISIykOKyCpy8md2kepH6ClnXx1hXe/ifq7JB0nXVUVrKmlFRURHWrFmDHTt2QKPRoHv37pg5c6ZZj4nujsEIEZGBnL6ZrYYsAr3dENXGq8nPn96nraofOX0rB1fT8q1ucTx9pjIb061bt9SwzMWLF+Hk5IQpU6Zg3rx5qqsqWTYGI0REBp5RIh/KzZmlIUHM2C5trGrxPI1GaxHFq/Hx8fjkk0+Qk5OD1q1b44knnsDgwYM5W8ZKMBghIjJwf5GhTawXqWl2/zD19buTSeqD3tJdSs1TU5Kli2zPtr5mO46wsDC0a9dOFaguXboUoaGVjeTIOjib+wCIiGyBDM9IG/i7LY53N/d0D4KPuzMSs4tUpmVYVPP3ZQq6rMiAiNZwdnI0+WyZ4OBgNSQjlwULFsDV1ZXZECvEzAgRkQGcTcxWM0r8vVzROaj5M0rcXZxwX+9Qqylk1dWLmHKIRmbLHDhwQK0tU3OmjJubGwMRK8VghIjIAI7E/bJibUu7euqGaraeTVYdXS2VBAWmDkYKCgrw5Zdf4scff1SvX1hYaPWrHROHaYiIDFu82oJ6EZ2BEa0R7u+Bm7eLsPN8Cu7vWznl19Ik3C5EWl4JXJwc0K99K6O/3o0bN/Dtt98iPz8fzs7OmDp1Kvr27ctsiL1mRlasWKHa6sp0qSFDhuDYsWONbv/WW2+ha9euquudtOB94YUX1BoBRES2oLxCgxNVtRPN6S9Sl2RWZvWrzI6st+D28LqsSO+wVmp4yVikX8i+ffvw+eefq0CkTZs2WLx4Mfr168dAxF6DkW+++QYvvvgili1bhpiYGPTp0weTJ09GWlr9izt99dVX+N3vfqe2v3DhguqIJ/v4/e9/b4jjJyIyu9ikXBSUVsDPwwXdQnwMss9ZVe3hD1xJR1pusUUHI8ZuAZ+Xl4dDhw6p4RjJhDz55JMICgoy6muShQcjb775popIZcXDHj16YOXKlWr55Y8//rje7eUNJCsjyjLNkk2ZNGkSHnroobtmU4iIrMWRqhbwgwzYgTQy0Av927eCzO7daKHt4XUzaaROxpj8/Pxw//33q06q8lVmzJAdT+0tLS3FiRMn8NJLL1Xf5+joiAkTJuDw4cP1Pmf48OFYtWqVCj6kAU1cXBy2bt2KhQsXNvg6JSUl6qKTm5urvpaVlamLoej2Zch9Es+zOfE9bZ7zfPha5UJsgyL8DPr75P4+oYhJyMa6E7fw6FDLWmVWakVuZBZCanV7t/U26PctwzIyW6Zt27bqtuy7c+fO1dfJen5v6LvPJgUjsvJhRUWFmtddk9yW9rv1kYyIPG/kyJEqxVZeXq6Wb25smOb111/H8uXL77h/586dKgtjaLt27TL4Ponn2Zz4njbdeZbMxdFrUi/hgLLE89i69bzB9u9SBjg5OOFiSh4+WLsV7ZreYd5oTmZIBsgJbT20OLjHcL9D5Y9e6aYqs2akSFXWluH72TSMcZ5ltpNFND3bu3cv/vKXv+Cdd95Rxa5Xr17F888/j9deew1/+MMf6n2OZF6kLqVmZkQKX2WIx9fX16ARm5z8iRMnwsXFxWD7JZ5nc+F72vTn+VJaEYqPHIG3mzOenDtRrS1jSHsLT2Hn+TRk+kRh8b1dYSmOb74A4CbG947A1KndDLLPa9euYePGjWqxOxmKkd/5st4Mf0db7+8N3ciGQYORwMBA1eVOlmSuSW6HhITU+xwJOGRIRgqORK9evVTEu2TJErz88stqmKcuaVwjl7rkJBkjaDDWfonn2Vz4njbdeT5xM0VdH9ShNdzdDF/LMHdAuApGNp5JwUtTe5i8y2lDfo6vXJ14WFRgi39/SsZ9z549+Omnn9Rt+TyZO3eu+uNTghG+n03DGOdZ3/016V0tkeqAAQOwe/fuWmN7cnvYsGENpmjqBhwS0Ag2qiEiW2l21pIW8I0Z2zUIrT1dkJ5Xgp+uVRbKmltOUZlak8YQzc5kWOazzz6rDkQGDRqkFrkLCLDsNvhkWE0OsWX45IMPPlBvHpmq+/TTT6tMh8yuEYsWLapV4Dp9+nS8++67WL16Na5fv65SQZItkft1QQkRkTWquWLtUCMFI67Ojpjep7KQc4OFtIc/EX8b0vRUZvy08bkzi93Uv5xllV3Jhs+bN081MpNaEbIvTf6Jz58/H+np6Xj11VeRkpKi5nxv3769uqg1ISGhVibklVdeUU1p5GtiYqJqViOByJ///GfDfidERCZ2KTVfZQm8XJ0QbcQVa6U9/OeH47H9XAryS8pVfYoldJtt7pReGZaROgVpnCmfD9OmTcPYsWNVUEL2qVnv6Oeee05dGipYrfUCzs6q4ZlciIhsyTHdirUd/I1ay9EnzA8dA70Ql1GAbWeTMW+geaf5/qxbj6YZzc6ysrJUS3cvLy88+OCDKhiREgD2DrFvllEJRURkhY7dyFJfhxi5A6l8YM/uX9mRdcNJ87aHl4X7ztzKadb3LUP77733nsqSSxZdAhMiwWCEiKgZpGbi56pgZKgBFse7m5lV7eEPx2UiMbsI5nLyZhbKNVqE+LojrLWHXs+R/lLbtm3DmjVrVEPLsLAwLF26FP7+plnplywfgxEiomZIKQKyCsvg7uKIXu2Mv2JtWGtPlYmQIOg7M2ZHfr6eVT1EIxmbu7l9+7ZaLkS3BIh05X7sscfQqpXxzxlZDwYjRETNcDW38oN4QERrNePFFOb0D6seqjFXa4RjNzL1XhxPjlGyIcnJyWrVdlmXTBprcSYl1cVghIioBcHIkEjT9cOY0isEbs6OuJqWj6+OJuD7U4k4fC0TFdKT3gTKKjSIqWp2ps9MGsmc3HfffYiMjFTLgHTp0sUER0nWiJO5iYiaSP7i/yUYMV3dg4+7C3q188Px+Cy8/F1s9f2hfu5YNr0H7o0ONerrxybmoKisAn4eLugc5F3vNpmZmaort6zqLqQ+RLpw6zOkQ/aLwQgRURPFZRQiv8xBDc/0CTdd7cP22GQViNSVklOMp1fF4N1H+hs1INE1eJOuq471rMFz5swZbN68WQVr0kFV13+KgQjdDYdpiIia2V+kb5gf3F1M00lahmKWb6p/RWDdII08bswhm2NVxauDI2s3J5MGZrLA3YYNG9R1yYYYY4V1sl3MjBARNfdDuYPpOoYeu34byTnFDT4uIYg8LtsNiwowauv7wTXqZKQj99q1a9VXMWbMGIwePbreRVCJGsJghIioCWQIQtdfpG6GwJjS8ooNul1TXUmrbH3v4eKEnlWt70+fPo0tW7aobIi3tzdmz56tilWJmorBCBFRE8RnFiI1rwRODlr0DTNdvUiQj7tBt2vu0FT/iFZwqWp9n52drQKRjh07YtasWSogIWoOBiNERE1w9Hpln40Ib8DD1XQrj0tfD5k1I8Wq9VWFSDlpiJ+7Xv0/mkOGf8SgiF/2P2rUKNW8rHfv3ixSpRbhoB4RURMcjav8UO7ka9qmY06ODmr6rqhvkqwczav39VDbGWVoKi4TXZzSgcs/qmyIkLqQPn36MBChFmMwQkTUBEerMgRRJg5GhEzblem7kgGpj7HWrIlLyUGX4gsY4RqPnIxUnDx50iivQ/aLwzRERHq6ebtQfeA7Ozog0sc87dglIJnYI0QNm0ixqtSIXEzJVdN6/7L1AnqE+mJ4p0CDvZ60cl+zajWinHOhhQMm3DMegwYNMtj+iQSDESKiJmZFotv5ws2psnbEHGQopub0XVk1ODYxF9/G3MKzX8Vg03+MVAvrtXRo5vjx49ixYwcqKiqQr3FFq54jMXLkSAN8B0S1cZiGiEhPR+IyTd5fRB/S4fTPs6JVq3hZSfipVSdQXFbRon3u378fW7duVYFIhlMANpb0wLDeXFuGjIPBCBFRE2fSWFowIqQT7MqFA+Dv5aqyJL/fcLZFK/v27dsXXl5eGD5mHDbld0CpgzMG1JhJQ2RIDEaIiPSQlF2Em7eLIJNV+re3vGBEtGvlgX8/3E8N46yPScRnh27o/VwJXOLj46tv+/n54de//jXKAjqr+Ttdg33UAnlExsBghIioCVmR6HZ+8HG33HK74VGBeGlKN3X9f7ZcwNGqoaXGFBUVYc2aNfj0009x+fLl6vtdXV2r+4uYcnVisj8MRoiImtBfxBo+lJ8YGYn7+7ZFuUarClqTcxqe8nvr1i289957uHjxIpycnJCfn19/szMr+L7JejEYISJqwkyaITUWibNUUtD619m90T3UFxn5pXhqVcwdBa0yLHPo0CF88sknyMnJQevWrfGrX/0K/fv3r95G1qK5kJKrrg/uwGCEjIfBCBHRXaTlFuN6RgEcHKwnQyCt6t9fOACtPF1w+mY2ln1/rrqgtbCwEF9//TV27doFjUaDnj17YsmSJWjbtm2tfcTEZ0Ge0iHAE0G+xlnzhkgwGCEiuosjVVkRaShmTUWc4f6e+NdD/VTR7TfHb+KrYwnq/hs3buDKlStqWGbatGmYM2cO3N3dG1wcbxCzImRklluFRURkIXRFoNYwRFPXqM5t8JvJ3fC/2y/ijxvPoVuIDwb06IGxY8eia9euCAkJafC5unoRYy2+R6TDzAgRkb71Ih2t80N54cBgPBiUDKeKyvoRGXYaM2ZMo4GI1JicuZWtrjMYIWNjMEJE1IiM/BJcTcu32iJOGZKR2TIeeYmY7HML6XklePrLGJSWaxp93qmb2Sir0CLY1w3t/VvWWp7obhiMEBE1QjdUIcMbrb1creZcSWHqvn378Pnnn6vpuoGBgXhs7n2qR8qJ+Cz8afO5Rp9fPaW3g7+anUNkTKwZISLSYz0aa+gvoiPBx/r163H9+vXq1u5TpkxRTczeftADT3x2HKuOJKB3u1Z4YFB4vfv4uap4lUM0ZArMjBAR6dPsrKN1FK+mpKRg5cqVKhBxcXHBzJkzcf/996tARIzvFowXJ1QuePfKd7FqOKau8gqNyp4IBiNkCgxGiIgacLugFJdS86zqQ1mal8k03aCgINU7pE+fPnds8+y4TpjYIxilFRo89cUJVUdS07mkXBSWVqhpzF2CfEx49GSvGIwQEd2lbqJTkDcCvd0s9jwVFBRUNzRzc3PDggUL8OSTT6o6kfo4OjrgzQf6oGMbL6TkFquW8WUVmjuGaAZGtFbbEhkbgxEiorssjmfJ9SJXr17FO++8g6NHj9bKjsgQTWN83F3w/sKB8HZzVkHXn7dcQIVGi8PXMvHdqUS1zcAOlrk6MdkeFrASEVlhvUhFRQX27NmDn376Sd2OjY3F4MGD4eio/9+YkvGRDMmSL07g00M3sOFkolqPRufDA9cRGeiFe6NDjfI9EOkwM0JEVI+cwl8WiRtqYZkRWdjus88+qw5EBg4ciMcee6xJgYjOpJ4hmBpd2fysZiCiq5l5elUMtscmG+jIierHzAgRUT2kbkLKMDoGelnUInGXLl3C999/j6KiIlUfMn36dLXQXXPJ0ExMwp0zaoRUoUjFyPJN5zGxRwicWD9CRsJghIiosXoRC2oBLxmRNWvWqIZmoaGhmDt3Lvz9W3Z8Ui8iRawNkYAkOadYbTcsyvKGq8g2MBghImpsPRoLWhzPz88P99xzjwpKJk6cCGfnlv8KT8srNuh2RM3BYISIqI684jLEJuZYRGbk4sWLanZMcHCwuj18+HCD7j/Ix92g2xE1BwtYiYjqOB6fBY0WaoG4UD8Ps5yf8vJybNu2Dd988w3Wrl2L0tJSo7yONHML9XNXtSH1kfvlcWtp+kbWicEIEVFDU3rN9AF8+/ZtfPzxxzh27Ji63aVLFzg5ORnltaQoddn0Hup63YBEd1seZ/EqGROHaYiIGloczwz9Rc6dO4dNmzahpKQEHh4eam0ZCUaMSfqIvPtIfzVrRopVdUL83FUgwj4jZGwMRoiIaigoKcdZXb2ICTMj0sRs+/btOH78uLodHh6OOXPmqKJVU5CAQ6bvyqwZKVaVGhEZmmFGhEyBwQgRUQ2yWq303mjXygPh/p4mOzcODg7IzKzMyIwcORLjxo1rVhOzlpDAg9N3yRwYjBARmXE9GlngTgIRCTxmzZqFtLQ0REVF8WdCdoXBCBFRvevRGDcYKSsrU7NlpDB12rRp6j4fHx91IbI3DEaIiKraoh+8ko6TCVnqfAyMMF4wkp6ejnXr1qksiJAF7tq0acOfA9ktBiNEZPdkIbi6M0kWfHQUfzTCTJJTp05h69atKjPi5eWF2bNnMxAhu8dghIjsmgQisjKtrMFSU2pOsbpfprwaIiCRpmUShJw+fVrdjoyMVIGIt7d3i/dNZO0YjBCRXQ/NSEakbiBi6BVrpUh11apVuHnzpipWHTNmDEaNGmXy2TJElor/E4jIbklPjZpDM42tWNsSEoDImjJSnLpo0SIVjDAQIfoFMyNEZLeMuWKtdFCVviFt27ZVt7t166am7Lq4uDR5X0S2jpkRIrJbxlqxNiUlBR988IEamsnNza2+n4EIUf2YGSEiu6VbsbahoRqHqvVZ9F2xVmpDpJ37jh07VHt3X19fFBQUqK9E1DBmRojIbklR6h+mVa5YixauWFtcXKx6h8iMGQlEZHG7pUuXIjTUsFODiWwRMyNEZNfKNBr1VcKNmrNqmrJibVJSkgpEsrKyVGHqhAkTMHToUFW4SkRGyoysWLECHTp0gLu7O4YMGYJjx441un12djaeffZZ9ReCm5ub+otB/nogIjL31N7/231FXf/PCZ3x9eKhePvBvurrwd+O17u/SExMjApEZIXdxx9/HMOGDWMgQmTMzMg333yDF198EStXrlSByFtvvYXJkyfj0qVLCAoKqrfRz8SJE9Vj8pdDu3btEB8fj1atWjX1pYmIDGrzmSRcSy+An4cLfjUyEj7uzZvpIr8DpTh19OjR8PDw4E+JyNjByJtvvonFixer6F9IULJlyxZ8/PHH+N3vfnfH9nL/7du3cejQoepKcsmqEBGZOyvydlVWZPGopgUiiYmJSEhIUAWrQn63SUBCRCYIRiTLceLECbz00kvV9+nGRw8fPlzvczZu3KhSljJM8/3336s1GB5++GH89re/VatVNjQ/Xy46uqlxspaDXAxFty9D7pN4ns2J72n9fX86GXHpBWjl4YKHB4Xp9XtAgg8Zlt6zZw80Gg2OHj2qakPIOPh+tv7zrO8+mxSMZGRkqCrx4ODgWvfL7YsXL9b7nLi4OPz4449YsGCBqhO5evUqnnnmGXWAy5Ytq/c5r7/+OpYvX37H/Tt37oSnpycMbdeuXQbfJ/E8mxPf042r0AJ/OyV/DDlgRGAxDvy4867ntLy8XGVDdH8cyVCzrL7L+jfj4/vZes9zYWGhZcymkb8epF7k/fffV5mQAQMGqBTnG2+80WAwIpkXqUvRkf/84eHhmDRpkkHn60tAJCdfalrYjMh4eJ5Nh+daP9+fSkLakViVFXlt0Xj4uDf+q1DWlJHMrvwukt9j48aNU4GI/E7i7w7j4fvZ+s9zzaZ/BgtGAgMD1X/E1NTUWvfL7ZCQkHqfIzNo5JurOSTTvXt31aFQhn1cXV3veI7MuJFLXbIfY/zHN9Z+iefZXPieblh5hQYr9l1X1xeP7gh/H4+7zpTZvHmzGqLx9/fHvHnzEBAQoDIiPM+mwfNsvedZ3/01aWqvBA6S2di9e3etzIfclrqQ+owYMUINzch2OpcvX1ZBSn2BCBGRMW08nYTrGQVo7emCR4ffvZhe1paR2rhevXphyZIlDf7hRUQm7DMiwyey5sJnn32GCxcu4Omnn1btjnWza2RFypoFrvK4zKZ5/vnnVRAiM2/+8pe/qIJWIiJTZ0V0fUUkK+LtVn9yOD8/v/q6BB9PPfUUZs2aVW/Gloharsk1I/Pnz1djpa+++qoaaunbty+2b99eXdQqBV41l8aWWg9Zp+GFF15A7969VZ8RCUxkNg0RkalrRW5kFqqsyKJhd2ZFJIN78OBBHDhwQP2BpVtxV4aoich4mlXA+txzz6lLffbu3XvHfTKEc+TIkea8FBGRwbIi//qxMiuyZHTUHVkRyYasX78e169X1pNI5lcXjBCRcXFtGiKyC99VZUX8vVyxaFjEHS0IJBCRIWcpuJs6darK+hKRaTAYISI7y4p0hFdVVkSGZfbt24f9+/er29KGYO7cuao5IxGZDoMRIrJ5G04mIr4qK7Jw6C9ZkdjY2OpApF+/fpgyZQqn+ROZAYMRIjtck+XY9dtIyytGkI87Bkf6w8nRwaazIv/ec1VdX1ojKyJkuq7M8uvatau6TkTmwWCEyI5sj03G8k3nkZxTXH1fqJ87lk3vgXujQ2GL1ldlRQK8XLFgSLhatHPgwIGqz5GDg4MaliEiK+szQkTWG4g8vSqmViAiUnKK1f3yuK0pk6zIj5VZkSeHBGPNV6tU22uuJ0NkWRiMENnJ0IxkRCoXvK9Nd588LtvZkg0xiUi4XYgeXvnIPbldrTEjGZHOnTub+9CIqAYO0xDZAakRqZsRqUlCEHlcthsWFQDbyYpcwkDnm+ilSUVxceVaWTIsI2vMEJHlYDBCZAekWNWQ21mDb366iF6FpxDkUqBuDx48WK1K6uzMX3tElob/K4nsgMyaMeR21pAV+fhQAgY7lMDB2RXzZs9Uq4UTkWViMEJkB2T6rsyaaWioRib2hvhVTvO1ZtLETNbG+vbELcRlV8DJqzu+WDwGIUG2MfREZKtYwEpkB6SPiEzfbYw8bs39RmR18I8++ghnzsZW9xWZP7YvAxEiK8DMCJGdGBIZAFdnR5SWa2rd7+bsiLcf7GvVfUbOnTuHTZs2oaSkBJu37URiVmcEentgwZDaa9AQkWViMEJkJz4/HK8Cke4hPnh1eg/EJGTjjR2XIMmQ8d2CYY3Ky8uxY8cOHD9+XN1uFxaGr9LaQgsNnh4bBQ9XJ3MfIhHpgcM0RHagsLQcnx66rq4/M64ThkUF4ukxUQj0dkVRmQbH42/D2mRmZqphGV0gMmLECHj0uAdxORq08XHDgiHtzX2IRKQnBiNEdmDNzzeRVViGiABPTIkOUfc5OjpgVOfK1WkPXMmANcnLy8P777+PlJQUeHp6YsGCBRg9djze2RenHpdAy92FWREia8FghMjGyTTXDw5UZkWWjO4IZ6df/tuP7hKovu6/nA5r4uPjg759+yIiIgJLly5Fp06dsPbETSRmFyHIxw0PMytCZFVYM0Jk4zadTlIf0oHebpjTP6zWYyM7VWZGziXlIj2vRA1vWKqMjAzVyt3X11fdnjRpklroTqbySi3Miqo1aKRWhFkRIuvCzAiRDdNotFi575q6/sTIyDs+pCX46BFa+eF+8KrlZkdOnz6thmW+/fZb1UtEODk5qUBErDl+E0k5xSor8tBg1ooQWRsGI0Q27MeLabicmg8fN2csGFr/h/ToLlV1I5ctr26ktLQU33//Pb777juUlZWpAETuq6mkvALvVPUVeYZZESKrxGEaIhul1Wrxzt7KD+lHhkXA192l3u2kbkSyJ/uvZKhMihS2WoK0tDSsXbtWDc/IcMyYMWMwatSo6myIzprjt1RWJNjXDQ8yK0JklRiMENmon29kqV4i0ujs8REdGtxuQERreLo6ISO/BBdSctGzrR/MHUSdPHkS27ZtU31EvL29MWfOHHTo8Mv3UKHRqhWGk7IL8c9dl9R9z4ztxFoRIivFYITIRr1blRWZOyCs0QXw3JydMLRjgBrSkSm+5g5GKioqcOTIERWIREVFYdasWfDy8qp+fHtsMpZvOl9rnR1J5rT2rD/zQ0SWjzUjRDboYkoe9lxKVx/SS0Z1vOv2oztbzhRfZ2dnzJs3D/fcc4/qH1I3EHl6VcwdC/5ptMDzq0+px4nI+jAYIbJB71f1FZnaKxQdAn/5MG+Iroj1+I0s1a3V1MMy0kX10KFD1fe1adMGI0eOVLUiNYdmJCOibWRf8rhsR0TWhcEIkY3JKAa2nE1R158aE6XXcyIDvRDW2gOlFRocicuEqcjCdjJdd8uWLfjhhx+Qmpra4LZSI1I3I1KThCDyuGxHRNaFwQiRjdmT5KiGLSTbEd1Ov/oPyUDoWsPvN9EU36SkJLz33ntqxV2ZITNhwgQEBQU1uH1aXsOBSHO2IyLLwQJWIhsiM2KOpjlUr8/SFGO6BOLrYwnYfyXd6MMyx44dw65du1Sxqp+fH+bOnYuwsNrdYetqrAi3OdsRkeVgMEJkQz4/nIAyrQP6hPlhaEf/Jj13eKdAODk6IC69ALeyChHW2tMogcj69esRGxurbnft2hX3338/PDw87vrcwZH+CPVzb3CoRkKwED93tR0RWRcO0xDZiLziMnx57Ka6vnRUZK3iT31IU7R+4a2MOlQjxxQeHq6GZSZPnoz58+frFYgICZR+P7Vb/fut+rpseg+1HRFZF2ZGiGyEDLHkFpcj2EOLe7pV1n80ldSNHI/PwoEr6QZb+VayIfn5+WqlXTFo0CDVPyQgIKDJ+5LvT0i8UXPSjGREJBC5NzrUIMdMRKbFYITIBsj6LB9WTee9p62m2S3dpTX8P3+4jINXM1BeoYGzU8uSp0VFRWpdmfT0dCxZsgTu7u4qO9KcQERW5n1nT+Wify9P644eoX6qWFVqRGRohhkRIuvFYITIBmyISURaXglCfN0wILCg2fvpHdYKrTxdkF1YhtO3sjEgovn1Fzdv3sS6deuQm5urFri7desWOnXq1Oz9rTtxC4nZRWql4QVDItj6nciGsGaEyMpJk6/39sep678a0QHOLfhfLdmFEZ0qu7Hua2bdiAzLHDx4EJ988okKRPz9/fHkk0+2KBCRrMiKqpV5ZZaQu4tTs/dFRJaHwQiRldtxLgXXMwrg5+GCBwa0a/H+dK3hpW6kqQoKCvDVV19h9+7dKiiJjo5WwzMhISEtOqb1Mb9kRQxVy0JEloPDNERWTD7w391bWUfx6PAO8HJr+X9pXWv40zezkVNYBr8mLEAnvUOuXr2q1pe599570b9//ybP6qmrrEKDf1dlRZaO7sisCJENYjBCZMV+upqJs4k5cHdxxGPDOxhkn6F+Hugc5I0rafmqkHVab/1nqEycOFENzci03eDgYIPVw9zKKkKgd2WtCBHZHg7TEFmxd/dVZgweHNQe/l6uBtvvL63hGx+qkSm7R44cqb4tK+wuWrTIYIFI3ayIhytrRYhsETMjRFZKhlEkM+Ls6IAnR0UadN8yxffjn66ruhEZCqpvqOX69euqm6oEJNK4rE+fPjC0704mIuF2IQK8XLFgKGtFiGwVgxEiK7VyX2WtyIy+bQ3eun1IZABcnR2RlFOMa+n56BRU2bBMaDQa7Nu3D/v371e327Rpg7Zt28LQymtkRZaM7ghPV/66IrJV/N9NZIUkQNh+LkVdf6qJC+LpQ4ZDhkT648CVDDXFVxeM5OXlqWzIjRs31O1+/fphypQpcHHRv8hVX9+fSkJ8ZqEaflo4jLUiRLaMNSNEVuj9fXHQaoEJ3YPRJfiXrIUhjaqa4qurG4mLi8PKlStVICLBx6xZszBjxgyjBCKSFfnXj1fU9cWjmBUhsnUMRoisTEpOMdafvKWuPz3W8FmRulN8j17PRHFZhRqeKSwsVMWpS5cuRe/evY322htPJ+FGZiFae7pgEbMiRDaPwzREVkYKS8sqtGo9lgERrY32Ol2DfRDi44qUvFIcv5GFkZ07qVV2ZZE7Y2RDanaU/fePlbUiT47qaJDeKURk2ZgZIbIi0oTsyyPxRs+KiCtXrmCyw2l4O5Rgf1U31m7duhk1EBGbTichLqNArZEjjdyIyPYxGCGyIl8cuYGC0gp0C/HB2KphFEOrqKjAzp078fXXX8OxrAB9nJPv2m/EYK+t0eL/atSKeDMrQmQXmP8kshJFpRX45Kcb1VmRlrZZr092dja+/fZbtcKu6N1vAD4/rEVFSh7ScosR5OsOY9p8Jglx6ZXr7LBWhMh+MDNCZCXWnriJzIJShLX2wLRe+rdo19fFixfx3nvvqUDEzc0NDzzwAGbNuA8921XWpey/0rxVfJuSFfmXrlZkZCR83I07HEREloPBCJEVkLbo7+2Lq26L7uxk2P+6Fy5cwDfffIPi4mLVwExmy3Tv3l09NlrP1vAttfVsMq6m5cPX3RmPjmCtCJE9YTBCZAW2nElGYnaRaos+b2C4wfffuXNnhIaGYujQofjVr36F1q1b39FvRBbN02i0MAbZ7//trqwVeWJkR/gyK0JkV1gzQmThZG2Yd/dWtn7/1chIuLsYZrE4aWLWoUMHODo6wtnZWQUh8rWu/hGtVSHp7YJSnEvKRa8wPxja1thktUqwj7szHmNWhMjuMDNC1IIah8PXMvH9qUT1VW4bY/+vb7uAS6l58HJ1wiNDW94Wvby8HFu3bsUXX3yBvXv3Vt9fXyAiXJwcMSwqQF3XTfE1VlbkVyMiVfEqEdkXZkaImmF7bDKWbzqP5Jzi6vtC/dyxbHoP3BsdapT9i8PXMlq0/8zMTKxbtw4pKZXr2khX1YZW5a3bjXXX+VTsu5yOZ8d1giHJGjuXU/Ph4+asghEisj/MjBA1I1B4elXMHYGCtGmX++VxY+xf+ou0ZP+xsbF4//33VSDi6emJhx9+GBMmTNBrivDoqrqRmPgs5JeUwxhZkcdHdICfJ7MiRPaIwQhRE4dOJGNR34CM7j55vLlDNo3tH83cf1lZGTZt2qT6h5SWlqJ9+/ZqtowUreorIsALEQGeKK8aOjKUnedTcDElT9WkSD0MEdknDtMQNcGx67fvyFjUJCGCPN73TztVrUVzpvDmFZffdf9yHLo6jrvJycnBmTNn1PVRo0Zh7Nixqmi1qWSK7xeZ8WqK78QewTBEVuTt3VersyKtPF1bvE8isqPMyIoVK1QVvru7O4YMGYJjx47p9bzVq1erlPDMmTOb87JEZpeW13AgUpMEFDL7pKmXxgKR5hyHCAwMxP33349HHnkE48ePb1YgUnMVX0MVse66kIoLybkqK/IEsyJEdq3JmRFpjPTiiy9i5cqVKhB56623MHnyZFy6dAlBQUENPu/GjRv4r//6L/WXGZG1CvLRrx36G3N7o094qybv//TNbPxm3ZkWHYesLSOzZfr166eGZER0dDRaamhHfzg7OiA+sxDxmQVq6Ka5pGhWVyvy6PAIZkWI7FyT/0R68803sXjxYjz++OPo0aOHCkqkGO7jjz9u9JfjggULsHz5cnTs2LGlx0xkFvIBGpOQ1eg2DlWzamb3D0OXYJ8mX+R58nyHu+x/cKR/vY+np6er1XZPnTqF9evXq/97hiLt2aXniCFaw/9wIU31LJHpyk+O5O8EInvXpMyIFL+dOHECL730UvV9kvKVivzDhw83+Lw//elPKmvyxBNP4MCBA3d9nZKSEnXRyc3NrS7Ek4uh6PZlyH2SbZ5nqeVYtukC1p5IrBUY1Cwj1QUQL0/pCk1FOTTNjAPk+f+x+nST9i+BktSF7NixQ/UR8fLywn333aem7srFUEZG+at6lX0XU/HggLbN2occ61s/XFLXHxnSHt6uDlb33rCF97Q14Hm2/vOs7z6bFIxkZGSov7SCg2sXr8ltWWSrPgcPHsRHH32k/lLT1+uvv66yKHXJsuaShTG0Xbt2GXyfZDvnubAc+OSyIy7nOMIBWszqoEErV2D9DUdkl/6Sw/Bz1WJ2Bw0q4k9ga3zLXvPxLg5671/+T8ridllZlVkbHx8fNTxz7tw5dTEkx3z51xkHrqRh0+ataM4SObG3HXAuyQmujlq0L7qCrVsrh2uskbW+p60Nz7P1nufCwkLzz6bJy8vDwoUL8cEHH6giOn1J5kXqUmpmRsLDwzFp0iT4+voaNGKTkz9x4kS4uLC/gbFY83m+mVWIxV+cxLWcAni6OuGfD/TG+K6VhZz/rdHieHwW0vJKEOTjhoERreHkePeeHfqYquf+CwoKVCdVCUSkOHzkyJHq/4v8XzHGuZYZMB9d24uswjKERA/DoA6/rGGjb1bkg5VH5X81Hh0eiQcmd4E1sub3tDXhebb+86wb2TBoMCIBhZOTE1JTU2vdL7dDQkLu2P7atWuqcHX69OnV9+lSxtJ6Wopeo6Ki7nieLF8ul7rkJBnjP76x9kvWfZ5PxGdhyefHkVlQihBfd3z02ED0bPvLuizynYzs0vIprg3RZ/9+fn4qMym/TObMmaNW3JXiVWOe61Gd22Dj6SQcisvC8M4NF63X58eLqYhNyoWHixOeGtvJqt4PtvCetlY8z9Z7nvXdX5OSrK6urhgwYAB2795dK7iQ28OGDbtj+27duuHs2bNqiEZ3mTFjBsaNG6euS7aDyBJtPpOEhz44ogKRnm198d2zI2oFIuZUs6ZKsiHyf+qpp55CRETL160x5hRfyYq8/UPlkMzCYREI8L7zDw4isk9NHqaR4ZNHH30UAwcOxODBg9XUXkkVy+wasWjRIrRr107VfUgfkrpTClu1amWwqYZEhiYfmO/svYY3dlQWWE7oHoS3H+wHLzfL6A+YnJys1pYJDQ1VmRAJRuT/mSnpWsOfTcxRvVH8vfRrVrb3cjpO38qBu4sjFo/iDBoi+kWTf8POnz9fTR989dVX1RoXffv2xfbt26uLWhMSEprdVInInErLNXh5w1msPXFL3ZZF216e1t1gdSAtDZJ+/vlnVcQtBatykT8CvL29TX4sQb7u6Bbio9q4H7yagRl92jYpK/LIkAi08WFWhIh+0aw/95577jl1qU/NJcnr8+mnnzbnJYmMKqewDE+tOoHDcZmQ2GP5jJ5YOKyDRZz14uJibNy4ERcuXFC3u3btqjqqenh4mO2YZKhGghFpDa9PMCJ9SU7dzFZZkSVjmBUhotosI/dMZCKywJz0yZB26tLFVJqH3coqxOOf/oy49ALVhOvfC/pjXNemFWYaS2JiohqWyc7OVhlHqXaXzsf6rLRrTLJOzfv743DgSrrKejR2PJVZkcvq+oIhEXp3sSUi+8FghOzG9thkteJtzYXuArxcUVKuQX5JOdr6yYyZQegearjp4y0hQzG6QERqrebOnavqsSzBwA6tVZYjNbcEl1Pz0TXEp8FtZSgnJiEbbs6OWMqsCBHVg8EI2U0g8vSqmFodTYXMlhHt/T2x7qlhqh7CUsg0ehmOOX78uOqmaupC1ca4uzhhSGQA9l1OV0M1DQUjNWtFHh7SnlkRIqoXK03JLoZmJCNSNxCpW7xqCVNNb968WV0bImR1bMmIWFIg0pQpvj9dzVSN21ydHfHUmDt7ChERCWZGyOZJjUjNoZn6pOQWq+2GRQWY7LjqZhAOHTqkevZIQ8A2bdo0qWuxOYzpEojXABy9fhtFpRXwcHW6Myuyu7JW5OHB7RFsQVknIrIszIyQzZNiVUNuZ4y1G7766iv88MMP6gNcZsvI+jKWLqqNt6qzkazSsRu373j88LVM/HwjC65OzIoQUeMYjJDN03f2hjlmecTHx2PlypW4evWqyohIbcjs2bPrXQ7B0sgMGmkNL6RupK63dlfWijw4OBwhfsyKEFHDOExDNk+m78qHYUoDQzUyKVUel+1M6cCBA9izZ4/KhgQEBGDevHl3rIht6aRu5JvjN+8IRiQrIsNekhV5eixrRYioccyMkM2TDqpTou9cyFHoumMsm97D5J1WZX0ZCUR69+6NJUuWWF0gIkZ2ClRN4q6k5SMpu6j6fl2tyAODwhDqZ77mbERkHZgZIZsntSDrYxLVdW83Z9VTREcyIhKI3BsdapJjkYUldcslyIKRYWFhqkbE3E3MmsvP0wV9wlvhZEI2Dl7JwAODwnE0LhNH4m7DxckBT4/tZO5DJCIrwGCEbJpkHl7eEIucojJEt/PFuqeGqw/Omh1YTZERkSBk//79uHbtGh577DHVQ0QusrK1tZO6ETmn60/egpuLo+rMKh4YGI52rZgVIaK7YzBCNm3j6STsOp+q/kp/Y24f1azL1NN38/LysH79ety4cUPdvnjxInr27Alb4eZcGcxJNkQuOj3aWkYnWyKyfAxGyGal55Vg2cZz6vp/jO9sljbvkgmRQESm77q4uKjZMrYUiEhn27/vqKwPqeuVDbGq3b6phsCIyHoxGCGbHZ555buzyC4sQ49QX5PP6JBhGZkpc/DgQXVbilOlk6qlNzIzdGdbeXxijxCTFwcTkXVhMEI2afOZZOw4lwpnRwf8fV4fuDiZduLYtm3b1JoyYsCAAZg8ebLKjNhTZ1sJUuRxc3a2JSLrwGCEbE5Gfgle/T5WXX9ufCez1C4MGzYMly5dwqRJkxAdHQ1bZOmdbYnIejAYIZsjgUhWYZmqEXnGRFNLKyoqVIFqVFTlcJC/vz9+/etfq66qtsqSO9sSkXVh0zOyKVvOJGPr2ZSq4ZnearVYY8vJycGnn36KVatWIS6uclqrsOVARMi06FA/9+rGcXXJ/aFm6GxLRNaHwQjZjMz8EvyhanjmmXGd0LOtn9FfU4ZiZG2ZW7duqfVkysrKYC+kKFUaxgkHC+psS0TWx7b/dCO78urGc7hdUIpuIT54bpxxh2dkWEZW2T1y5Ii63bZtWzVbpnXr1rAnMm333Uf6q1kzNYtZTd3ZloisG4MRsglbzyarIRqnqtkzxhyeycrKwrp165CUlKRuDx06FBMmTFAdVe2RBBwyfVdmzZi6sy0R2QYGI2T1JBvyh++qhmfGRiG6nXGHZ+Lj41Ug4u7ujpkzZ6q1ZeydBB6cvktEzcVghKyedFnNLChF12AfNZXX2Pr06aNavMtqu35+xq9LISKydSxgJatvR77pdJL6y/yNeb3h5mz4oZLbt29j9erVqqW7kBV2R40axUCEiMhAGIyQ1coqKMUrVcMzT43piN5hrQz+GrGxsXjvvffUrJmdO3cafP9ERMRhGrJif9x0Dhn5pegc5I1f39PZoPuWKbrbt29HTEyMut2+fXuMHz/eoK9BRESVWDNCVmnHuRR8fyoJMmFDZs8YcngmIyNDzZZJTU1Vt2VIZuzYsXB0ZCKRiMgYGIyQ1ckuLMXLGyqHZ5aOiUKfcMMNz0hL96+++kplRry8vDBr1qzqFu9ERGQcDEbI6kiDLVkMr1OQN5438PBMcHAwPD09VfOy2bNnw8fHx6D7JyKiOzEYIauy63wqNpxMVMMzb8ztDXcXJ4OsLePr66tmyXh4eOCxxx5TtzksQ0RkGhwEJ6sanvn9hrPq+uLRHdGvfctar2u1Wpw8eRL//ve/qwtVRatWrRiIEBGZEIMRshp/2nwe6XkliGrjhRcmdGnRvkpLS/Hdd99h48aNKC8vx5UrV1RwQkREpsdhGrIKuy+kYn1M1fDMvD4tGp6RWTJr165FZmamGpoZN24cRo4cqa4TEZHpMRghi5dTWFY9PPPkqI7o38zhGcl8yHCM9A+RbIgUp86ZMwcREREGPmIiImoKBiNk8V7bch6puSXoGOiFFyc2f3gmPT0dW7ZsUUFJ586d1SJ3MnOGiIjMi8EIWbQ9F9Ow7sQtyAiKrD3TkuGZoKAg1bzMyckJw4cP57AMEZGFYDBCFiunqAwvra8cnnliRCQGRPg36fmSATl+/Dg6duyIgIAAdd/o0aONcqxERNR8DEbIYv15y3mk5BYjMtAL/29S1yY9t7i4GJs2bcL58+dVI7Mnn3wSzs58uxMRWSL+diaLUaHR4tj120jLK0ZSdhHWHK8cnvnb3N7wcNV/eCYxMVGtLZOdna36hfTt21cNzRARkWViMEIWYXtssmrznpxTXOv+cV2DMKiDv97DMkePHsWuXbug0WhU87K5c+eiXbt2RjpqIiIyBAYjZBGByNOrYqBtoIBVHr83OrTRfZSUlGDDhg24dOmSut29e3fMmDED7u7uRjpqIiIyFHZgJbMPzUhGpLHep/K4bNcYqQcpKChQwzFTp07FvHnzGIgQEVkJZkbIrKRGpO7QTE0Sgsjjst2wqIDaj2m16iJ1IRKEyJBMYWEhQkMbz6IQEZFlYWaEzEqKVZuznQQdX3/9NX744Yfq+/z8/BiIEBFZIWZGyKyCfNybvF18fDy+/fZb5OXl4fr16xg6dCh8fX2NeJRERGRMDEbIrAZH+iPUz73BoRpZui7Ez11tJ0MyBw8exJ49e9R1aWQmtSEMRIiIrBuDETIrJ0cHzB8Yjrd2X7njMd0ausum90BxUaGaLXPt2jV1X+/evTFt2jS4urqa+IiJiMjQGIyQWZVXaLD9XIq67unqhMLSiurHJCMigcikHsF45513kJmZqWbNyGwZaWTmIB3RiIjI6jEYIbP65vhNXEzJg5+HC378f2NwOTVfFatKjYgMzUjmRIwZMwYHDhxQwzJt2rThT42IyIYwGCGzLoT3j52X1fUXJnRGgLcbhnm7qdv5+flISU6q7p7aq1cv9OjRg23diYhsEIMRMpt//3gFtwtK0SnIGwuGRlTfHxcXh/Xr16thmKVLl8Lb21vdz/VliIhsE4MRMovrGQX49NANdf2Vad3h4uSo1pPZu3evGo4RstpuaWkpf0JERDaOwQiZxZ+3XEBZhRZju7bB2K5ByM3NVdkQ6SEiBgwYgMmTJ8PFxYU/ISIiG8dghEzu4JUM/HAhVRWnvjKtB65evaqm7UpXVZmqO336dERHR/MnQ0RkJxiMkMmn8r62+by6vnBohKoX2fDT2eo1ZWR9GX9/f/5UiIjsCIMRMqnVP9/EpdQ8tPJ0wX9O6Kzuk74hEoCMGDFC9REhIiL70qyF8lasWIEOHTqoJdqHDBmCY8eONbjtBx98gFGjRqF169bqMmHChEa3J9uVW1SGN3ddRrhjNhaGpqneIsLNzU31EWEgQkRkn5ocjHzzzTd48cUXsWzZMsTExKBPnz6q0DAtLa3e7WV2xEMPPaTWEzl8+DDCw8MxadIkJCYmGuL4yYr8+8eriCq9hgluV5GfdA1nzpwx9yEREZE1BiNvvvkmFi9ejMcff1w1oVq5ciU8PT3x8ccf17v9l19+iWeeeUa17+7WrRs+/PBDNYVz9+7dhjh+shJJOSXIOLkT0c6p6rastMsiVSIiEk0aoJeeDydOnMBLL71UfZ+jo6MaepGshz6kULGsrKzRIsWSkhJ10ZFpn0KeJxdD0e3LkPukO507dw5J1y8j0LECFY4umD/7fnTp0kUFpXIhw+F72jR4nnmebUmZET8L9d1nk4KRjIwMVFRUqGZUNcntixcv6rWP3/72t2jbtq0KYBry+uuvY/ny5Xfcv3PnTpWFMbRdu3YZfJ9UKSUlRV3kjZam8ULPjhFqKq9cyHj4njYNnmeeZ1uyywifhZKA0IdJpy789a9/xerVq1UdiRS/NkQyL1KXUjMzoqs18fX1NWjEJid/4sSJbK5lJAkJN/H5qlU4VxaM7v0HY/F09g8xJr6nTYPnmefZlpQZ8bNQN7Jh0GAkMDBQrQ+Smlo57q8jt0NCQhp97t///ncVjPzwww/o3bt3o9vK7Aq51CUnyRgdOY21X3uVnZ2NVq1aqeuH0p2wrjgaWidXrJzQlefZRPie5nm2JXw/W+951nd/TSpgle6Y0qa7ZvGprhh12LBhDT7vb3/7G1577TVs374dAwcObMpLkpVF15s3b1ZTvyVAzSksw5s7L6FA64ap4ZrqqbxEREQtGqaR4ZNHH31UBRWDBw/GW2+9hYKCAjW7RixatEgt+y51H+J///d/8eqrr+Krr75SvUmkfkDISqy61VjJ+kk90bp166qzZrLGzI60TGQVlqFTGy8MD84x9yESEZGtBCPz589Henq6CjAksJApu5Lx0BW1JiQkqBk2Ou+++66ahSNtvmuSPiV//OMfDfE9kJlJvxDJiEhmxMvLC7NmzYKDbzA++26/evz3U7si7zIb3RERkQELWJ977jl1qY8Up9Z040blMvFkeyT42LZtG06ePKluS+Zr9uzZ8PHxwROf/oxyjRbjuwVhVKdAbL1s7qMlIiJLxYVAqNmkA68uEJF27qNHj1ZZsf2X07H7YhqcHR3w8rTuPMNERNQoBiPUbIMGDVJt/fv164fIyMjqVXn/Z0vlqryLhnVAVBtvNpUjIiLDL5RH9klqf2SNofLycnVbsiAyLKMLRMTXxxJwOTUfrT1d8Pw9lavyEhERNYaZEdKLzJKR2TIya6aoqAhTp069Yxs1lXdXZXHIixO7wM+TU3mJiOjuGIxQo7RaraoNkRlTkhGR4tSePXvWu+3bu6+oqbxdgr3x0OD2PLNERKQXBiPUIFmsUKbsxsbGqtudOnXCzJkz1fTduq6l5+Pzw5Uzp/5wXw84O3EEkIiI9MNghOqVlpaGb775Brdv34aDgwPuueceDB8+XF2vz5+3XFBTee+Rqbyd2/CsEhGR3hiMUIOt/2W1RVmYUBrWyUKFNVVotDh2/TbS8oqRnFOMHzmVl4iImonBCFWrqKhQCyEKWeju4YcfVosjenh41DpL22OTsXzTeRWE1DSmSxt0bMMW/0RE1DQc2CclKSlJLXB3+fIvrVIlG1JfIPL0qpg7AhEh2RF5nIiIqCkYjNg5mS1z5MgRfPTRR8jKysK+ffvUffWRoRnJiNT/aCV5XLYjIiLSF4dp7Jj0C9m4cSMuXryobnfv3h0zZsxosEhVakTqy4joSAgij8t2w6ICjHbcRERkWxiM2Klbt26pJmY5OTmqTmTSpEmqvXtDgYiQYlV96LsdERGRYDBih6SL6ieffAKNRoPWrVtj3rx5CA0Nvevzgnzc9dq/vtsREREJBiN2SGbI9OnTR601M336dLi5uen1PFkETxInDZSUQHIqIX7uGBzpb9gDJiIim8ZgxE4kJCQgICCgunvqtGnT1EJ3jQ3L6EhB6/v74/C/2y82GoiIZdN7wMnx7vskIiLS4WwaGyeBxIEDB/Dpp59iw4YN1TNlpE5En0Akv6Qcz34Vg9e3XYRMkpk7IAz/92BfhPrVHoqRjMi7j/THvdF3H+4hIiKqiZkRG1ZQUKACkGvXrqnbnp6earE7Fxf9VtOV9WaWfnECV9Py4eLkgGXTe2LBkPYqiJnWu211B1apEZGhGWZEiIioORiM2KgbN27g22+/RX5+PpydnTF16lT07dtXr2yI2B6bgv9ae1plRoJ93fDuIwPQv33r6scl8OD0XSIiMgQGIzZGZsjIsIyueVmbNm3U2jJBQUF6PV8alv1j5yW8s7cymyIZjxUP90cbH/2KXImIiJqKwYiNkWGYM2fOqEBEMiFTpkxRi97pI6ugFL9efRIHrmSo20+MjMTvpnSDixNLi4iIyHgYjNgYCTwkE5KWlqam7+orNjFH1YckZhfBw8UJ/zu3N2b0aWvUYyUiIhIMRmxgWGbv3r3w9vbG4MGD1X3SwEyfJmY6607cwssbzqKkXIMOAZ5YuXAAuoX4GvGoiYiIfsFgxIrl5uZi/fr1iI+PVz1DunTpglatWun9/NJyDf60+RxWHUlQt+/pFoQ35/eFn4d+s22IiIgMgcGIlbp69aqatltYWKiGZqSTalMCkZScYjz95QmcTMhWXVX/854u+I/xneDIhmVERGRiDEasTEVFBfbs2YOffvpJ3Q4JCVE1ItJdVV9H4zJVI7OM/FL4ujvj7Qf7YVw3/WbbEBERGRqDESsiM2S++OILNSwjZJVdWW1X+ojo+/yPf7qBv2y9oKbwdgvxwXsLByAioLJFPBERkTkwGLEi0rCsc+fOSElJwYwZM9CjR496t5NAo2531JLyCvzu27PYeDpJbTOzb1u8Prs3PFydTPxdEBER1cZgxAqGZaSLqp+fn7o9fPhw9OrVC76+9c922R6bjOWbziM5p7j6vjbebnB1dkBidjGcHR3wyrTueHR4B727sRIRERkTgxELlpWVpVq6l5SUYPHixapQVQKIxgKRp1fFoO7Cuun5Jeqr1Id89NggDOrgb4KjJyIi0g+DEQt14cIFfP/99yoQcXd3R3p6Otq1a9fg9jI0IxmRuoFITTIkU3N9GSIiIkvAYMQC27nv2rULx44dU7fDwsIwZ86cu07blRqRmkMz9UnNLVHbcYE7IiKyJAxGLMjt27exbt06JCcnV9eHjB8/Hk5O9ReZ5hSW4XBcJg5fy8D2cyl6vYYUtRIREVkSBiNmUt+Ml507d6pAxMPDAzNnzlQdVWsqLC3HzzeycOhaBg5dzURsUg60jY3L1ENei4iIyJIwGDGD+ma8hPq546UJA9HNwUGttCtFqtKu/dTN7Org4+TNLJRV1I4+OgV5Y3hUAIZG+mP55vNIyy2pt25E5s2E+FUGPURERJaEwYiJ1Zzx4utQjPZO2YgtD1Ht2Z//9hJ+N2UgvjqZjkPXLuLn67dRVFZR6/ntWnmo4GNEp0BV+xHs+0umQ1q5y74l8KgZkOgm8C6b3gNObPdOREQWhsGICdWc8dLRKRPDXeLh4qBBnsYN8ZrKWS6vb7tY6zkBXq4q6JDgQ4KQ9v6eDfYHuTc6FO8+0v+OrItkRCQQkceJiIgsDYMRE1I1IjkFGOFyE12cM9R9yRU+SNPUbsfev30r3Ne7LYZ3CkDXYJ8mNSeTgGNij5A76lGYESEiIkvFYMSEEpKSMd3tIlo7FqnC01PloThd3hba6oGUStId9f6+DfcUuRsJPDh9l4iIrAWDERM5e/YsLu/biNaO5SjUumB/aSSSNfV3UuWMFyIisicMRkzF0QmainIkVvhgf2lHFMPljk0444WIiOwRgxEjd1N1dnZGel4Jlu3LQkpJZ6RqfVEBB854ISIiqsJgxAi0Wi1iYmKwf/9+jJr2AJ5ffwmJ2UXwdQ/AZwsGIL+kjDNeiIiIqjAYMTBZ2G7z5s2IjY1Vt9/4cisSS9oiMtALHz06EB3beKv7OeOFiIioEoMRA0pJScHatWvVGjNSAXK8rC3OlodgZKdArHi4P/w8f6kT4YwXIiKiSgxGDDQsc/z4cezYsQMVFRXQOHtgW0EE0jTeWDQsAn+4rwdcnBwN8VJEREQ2h8GIAZw4cQJbt25V13NcA7ElJwzlji547f4eWDisgyFegoiIyGbxz3UD6NOnD1oFBOGiU0esz4mAm7s7Pnt8MAMRIiIiPTAYaeawzPnz59VXsedyJt5P64DD+f7oGOiN754dgZGdA5uzayIiIrvDYZomKioqwsaNG3Hx4kWMHz8eZytC8caOS6q9e32FqkRERNQ4BiNNcOvWLaxbtw45OTlwcnLClnPpWH0jVz3GQlUiIqLmYTCiBxmOOXz4MHbv3g2NRgNfv1Y47tAFP93QqCm6f5zOQlUiIqLmYjByF4WFhfj+++9x+fJldTusYxd8fisA8Tll8HV3xjsLBrA+hIiIqAUYjNyFDMlcu3ZNDcu07zMcf/25BIWlZegY6IUPa3RUJSIiouZhMFKPCo0Wx67fRlpeMYJ83HHf9OnYf7MUy35KY6EqERGRgTEYqWPT8Ths27oFR4uCkan1Uvd5uDiiqEyjrrNQlYiIyAL6jKxYsQIdOnSAu7s7hgwZgmPHjjW6vazX0q1bN7V9r169qruVWpo1e07g4ObVCNbexijXG1K6qu7XBSIPDgrHn+6PZmt3IiIicwYj33zzDV588UUsW7YMMTExqvvo5MmTkZaWVu/2hw4dwkMPPYQnnngCJ0+exMyZM9VFt6qtJZAZMnv27sX5/Zvh6VCGbI079pZ2VIvd1bTvcroawiEiIiIzBiNvvvkmFi9ejMcffxw9evTAypUr4enpiY8//rje7d9++23ce++9+M1vfoPu3bvjtddeQ//+/fHvf/8blqCsrAxff/019u/bp0KPK+UB2FTSHdlajzu2Tc4pVrUkREREZKaakdLSUrUo3EsvvVR9n6OjIyZMmKD6cNRH7pdMSk2SSfnuu+8afJ2SkhJ10cnNza0OHORiKOnp6bh06RLKy8vh4OSMfUVhuFbReBv35OwClJX5GuwY7IHuZ2bInx3xXJsT39M8z7akzIi/o/XdZ5OCkYyMDFRUVCA4OLjW/XJb2qPXJyUlpd7t5f6GvP7661i+fPkd9+/cuVNlYQzZzMzDw0OdLIeASFy7Wlmw2pi4c6ew9dZJgx2DPdm1a5e5D8Fu8FzzPNsSvp+t9zxLry6rnU0jmZea2RTJjISHh2PSpEnw9TVcVkKCkG3btqnMjpu7B9b/Yz9Sc0uqylZrkyGcED83PDd/tOq6Sk07z/ImnzhxIlxcuG6PMfFcmwbPM8+zLSkz4u9o3ciGQYORwMBA1fwrNTW11v1yOyQkpN7nyP1N2V64ubmpS11ykgx9opydnVW2Rfb7xxk98fSqGBV41AxIdKHHsuk94e7matDXtyfG+PkRz7U58T3N82xLXIzwO1rf/TWpgNXV1RUDBgxQa7TUnIkit4cNG1bvc+T+mtsLicAa2t6c7o0OxbuP9EeIn3ut++W23C+PExERkWE1eZhGhk8effRRDBw4EIMHD8Zbb72FgoICNbtGLFq0CO3atVN1H+L555/HmDFj8I9//APTpk3D6tWrcfz4cbz//vuwRBJwTOwRUqsD6+BIfw7NEBERWUowMn/+fDUL5dVXX1VFqH379sX27duri1QTEhLUDBud4cOH46uvvsIrr7yC3//+9+jcubOaSRMdHQ1LJTUhw6ICzH0YREREdqFZBazPPfecutRn7969d9w3b948dSEiIiIySDt4IiIiIkNhMEJERERmxWCEiIiIzIrBCBEREZkVgxEiIiIyKwYjREREZFYMRoiIiMisGIwQERGRWTEYISIiIuvrwGpqWq22SUsRN2XZ5MLCQrVfriZrPDzPpsNzzfNsS/h+tv7zrPvc1n2OW3UwkpeXp76Gh4eb+1CIiIioGZ/jfn5+DT7uoL1buGIBNBoNkpKS4OPjAwcHB4NGbBLg3Lx5E76+vgbbL/E8mwvf0zzPtoTvZ+s/zxJiSCDStm3bWovoWmVmRL6BsLAwo+1fTj6DEePjeTYdnmueZ1vC97N1n+fGMiI6LGAlIiIis2IwQkRERGZl18GIm5sbli1bpr4Sz7Mt4Hua59mW8P1sP+fZKgpYiYiIyHbZdWaEiIiIzI/BCBEREZkVgxEiIiIyKwYjREREZFY2H4ysWLECHTp0gLu7O4YMGYJjx441uv3atWvRrVs3tX2vXr2wdetWkx2rvZznDz74AKNGjULr1q3VZcKECXf9uVDzznVNq1evVh2MZ86cydNp4Pe0yM7OxrPPPovQ0FA1K6FLly78/WGE8/zWW2+ha9eu8PDwUF1DX3jhBRQXF/M93Yj9+/dj+vTpqguq/A747rvvcDd79+5F//791Xu5U6dO+PTTT2FUWhu2evVqraurq/bjjz/Wnjt3Trt48WJtq1attKmpqfVu/9NPP2mdnJy0f/vb37Tnz5/XvvLKK1oXFxft2bNnTX7stnyeH374Ye2KFSu0J0+e1F64cEH72GOPaf38/LS3bt0y+bHb+rnWuX79urZdu3baUaNGae+//36THa+9nOeSkhLtwIEDtVOnTtUePHhQne+9e/dqT506ZfJjt+Xz/OWXX2rd3NzUVznHO3bs0IaGhmpfeOEFkx+7Ndm6dav25Zdf1q5fv15mz2o3bNjQ6PZxcXFaT09P7Ysvvqg+C//1r3+pz8bt27cb7RhtOhgZPHiw9tlnn62+XVFRoW3btq329ddfr3f7Bx54QDtt2rRa9w0ZMkS7dOlSox+rPZ3nusrLy7U+Pj7azz77zIhHab/nWs7v8OHDtR9++KH20UcfZTBihPP87rvvajt27KgtLS1t2g/UzjX1PMu248ePr3WffGCOGDHC6MdqK6BHMPLf//3f2p49e9a6b/78+drJkycb7bhsdpimtLQUJ06cUEMANde4kduHDx+u9zlyf83txeTJkxvcnpp3nuuSpatlCWt/f3+eUgO/p8Wf/vQnBAUF4YknnuD5NdJ53rhxI4YNG6aGaYKDgxEdHY2//OUvqKio4Dk34HkePny4eo5uKCcuLk4NhU2dOpXn2YDM8VloFQvlNUdGRob6RSC/GGqS2xcvXqz3OSkpKfVuL/eT4c5zXb/97W/VWGbdNz+1/FwfPHgQH330EU6dOsXTacTzLB+KP/74IxYsWKA+HK9evYpnnnlGBdnS2ZIMc54ffvhh9byRI0eq1WDLy8vx1FNP4fe//z1PsQE19Fkoq/sWFRWpeh1Ds9nMCFmHv/71r6qwcsOGDaqAjQxHlu1euHChKhgODAzkqTUijUajsk/vv/8+BgwYgPnz5+Pll1/GypUred4NSIoqJeP0zjvvICYmBuvXr8eWLVvw2muv8TxbOZvNjMgvXycnJ6Smpta6X26HhITU+xy5vynbU/POs87f//53FYz88MMP6N27N0+ngd/T165dw40bN1QVfc0PTeHs7IxLly4hKiqK572F51nIDBoXFxf1PJ3u3burvzBlOMLV1ZXn2QDn+Q9/+IMKsJ988kl1W2Y8FhQUYMmSJSr4k2EearmGPgt9fX2NkhURNvuTk//88hfK7t27a/0iltsytlsfub/m9mLXrl0Nbk/NO8/ib3/7m/prZvv27Rg4cCBPpRHe0zJF/ezZs2qIRneZMWMGxo0bp67LtEhq+XkWI0aMUEMzumBPXL58WQUpDEQM837W1ZfVDTh0ASCXWTMcs3wWam182phMA/v000/V9KQlS5aoaWMpKSnq8YULF2p/97vf1Zra6+zsrP373/+uppwuW7aMU3uNcJ7/+te/qul869at0yYnJ1df8vLyDP8msPNzXRdn0xjnPCckJKgZYc8995z20qVL2s2bN2uDgoK0//M//9PCn7hta+p5lt/Jcp6//vprNf10586d2qioKDUTkhomv1ullYJc5GP/zTffVNfj4+PV43KO5VzXndr7m9/8Rn0WSisGTu1tIZkf3b59e/XhJ9PIjhw5Uv3YmDFj1C/nmtasWaPt0qWL2l6mNm3ZsqWlh2AXmnKeIyIi1H+Iuhf5RUOGPdd1MRgxzntaHDp0SLUCkA9Xmeb75z//WU2rJsOd57KyMu0f//hHFYC4u7trw8PDtc8884w2KyuLp7kRe/bsqfd3ru7cylc513Wf07dvX/VzkffzJ598ojUmB/nHeHkXIiIiIjutGSEiIiLrwGCEiIiIzIrBCBEREZkVgxEiIiIyKwYjREREZFYMRoiIiMisGIwQERGRWTEYISIiIrNiMEJERERmxWCEiIiIzIrBCBEREZkVgxEiIiKCOf1/YJotKDNHsygAAAAASUVORK5CYII=",
"text/plain": [
"<Figure size 640x480 with 1 Axes>"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"# calibration curve\n",
"from sklearn.calibration import calibration_curve\n",
"prob_true, prob_pred = calibration_curve(df_test[\"target\"], y_pred_proba[:, 1], n_bins=20)\n",
"plt.plot(prob_pred, prob_true, marker='o')\n",
"plt.plot([0, 1], [0, 1], linestyle='--', color='gray')\n",
"plt.grid()"
]
}
],
"metadata": {
"kernelspec": {
"display_name": "cat",
"language": "python",
"name": "python3"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.11.14"
}
},
"nbformat": 4,
"nbformat_minor": 5
}
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment