[Autogluon] (tripartite)_f1값이 너무 안 나오는데..(망함)

import

import pandas as pd
import os
import math
import numpy as np
import networkx as nx
import matplotlib.pyplot as plt
from autogluon.tabular import TabularDataset, TabularPredictor
from sklearn.model_selection import train_test_split
from node2vec import Node2Vec
from node2vec.edges import HadamardEmbedder, AverageEmbedder, WeightedL1Embedder, WeightedL2Embedder

# sklearn
import sklearn
from sklearn import metrics
from sklearn.metrics import precision_score, recall_score, f1_score

    

    def throw(df, fraud_rate):  # 사기 거래 비율에 맞춰 버려지는 함수!
        df1 = df[df['is_fraud'] == 1].copy()
        df0 = df[df['is_fraud'] == 0].copy()
        df0_downsample = (len(df1) * (1-fraud_rate)) / (len(df0) * fraud_rate)
        df0_down = df0.sample(frac=df0_downsample, random_state=42)
        df_p = pd.concat([df1, df0_down])
        return df_p
    
    def split_dataframe(data_frame, test_fraud_rate, test_rate=0.3):
        n = len(data_frame)
    
        # 사기 거래와 정상 거래를 분리
        fraud_data = data_frame[data_frame['is_fraud'] == 1]
        normal_data = data_frame[data_frame['is_fraud'] == 0]

        # 테스트 데이터 크기 계산
        test_samples = int(test_fraud_rate * (n * test_rate))
        remaining_test_samples = int(n * test_rate) - test_samples
    
        # 사기 거래 및 정상 거래에서 무작위로 테스트 데이터 추출
        test_fraud_data = fraud_data.sample(n=test_samples, replace=False)
        test_normal_data = normal_data.sample(n=remaining_test_samples, replace=False)

        # 테스트 데이터 합치기
        test_data = pd.concat([test_normal_data, test_fraud_data])

        # 훈련 데이터 생성
        train_data = data_frame[~data_frame.index.isin(test_data.index)]

        return train_data, test_data
    
    def concat(df_tr, df_tst):   
        df = pd.concat([df_tr, df_tst])
        train_mask = np.concatenate((np.full(len(df_tr), True), np.full(len(df_tst), False)))    # index꼬이는거 방지하기 위해서? ★ (이거,, 훔,,?(
        test_mask =  np.concatenate((np.full(len(df_tr), False), np.full(len(df_tst), True))) 
        mask = (train_mask, test_mask)
        return df, mask
        
    def evaluation(y, yhat):
        metrics = [sklearn.metrics.accuracy_score,
                   sklearn.metrics.precision_score,
                   sklearn.metrics.recall_score,
                   sklearn.metrics.f1_score,
                   sklearn.metrics.roc_auc_score]
        return pd.DataFrame({m.__name__:[m(y,yhat).round(6)] for m in metrics})
        
    def compute_time_difference(group):
        n = len(group)
        result = []
        for i in range(n):
            for j in range(n):
                time_difference = abs((group.iloc[i].trans_date_trans_time - group.iloc[j].trans_date_trans_time).total_seconds())
                result.append([group.iloc[i].name, group.iloc[j].name, time_difference])
        return result

    def edge_index_save(df, unique_col, theta, gamma):
        groups = df.groupby(unique_col)
        edge_index = np.array([item for sublist in (compute_time_difference(group) for _, group in groups) for item in sublist])
        edge_index = edge_index.astype(np.float64)
        filename = f"edge_index_attempt{self.save_attempt}_{str(unique_col).replace(' ', '').replace('_', '')}.npy"
        
        while os.path.exists(filename):
            self.save_attempt += 1
            filename = f"edge_index_attempt{self.save_attempt}_{str(unique_col).replace(' ', '').replace('_', '')}.npy"
        np.save(filename, edge_index)
        #tetha = edge_index_plust_itme[:,].mean()
    
        
        edge_index[:,2] = (np.exp(-edge_index[:,2]/(theta)) != 1)*(np.exp(-edge_index[:,2]/(theta))).tolist()
        edge_index = torch.tensor([(int(row[0]), int(row[1])) for row in edge_index if row[2] > gamma], dtype=torch.long).t()
        return edge_index
    
    def edge_index(df, unique_col, theta, gamma):
        groups = df.groupby(unique_col)
        edge_index = np.array([item for sublist in (compute_time_difference(group) for _, group in groups) for item in sublist])
        edge_index = edge_index.astype(np.float64)
       # filename = f"edge_index_attempt{self.save_attempt}_{str(unique_col).replace(' ', '').replace('_', '')}.npy"
        
        # while os.path.exists(filename):
        #     self.save_attempt += 1
        #     filename = f"edge_index_attempt{self.save_attempt}_{str(unique_col).replace(' ', '').replace('_', '')}.npy"
        # np.save(filename, edge_index)
        #tetha = edge_index_plust_itme[:,].mean()
    
        
        edge_index[:,2] = (np.exp(-edge_index[:,2]/(theta)) != 1)*(np.exp(-edge_index[:,2]/(theta))).tolist()
        edge_index = torch.tensor([(int(row[0]), int(row[1])) for row in edge_index if row[2] > gamma], dtype=torch.long).t()
        return edge_index

df = pd.read_csv("~/Desktop/fraudTrain.csv")

삼분그래프

def build_graph_tripartite(df_input, graph_type=nx.Graph()):
    df=df_input.copy()
    mapping={x:node_id for node_id, x in enumerate(set(df.index.values.tolist() + 
                                                       df["cc_num"].values.tolist() +
                                                       df["merchant"].values.tolist()))}
    df["in_node"]= df["cc_num"].apply(lambda x: mapping[x])
    df["out_node"]=df["merchant"].apply(lambda x:mapping[x])
    
        
    G=nx.from_edgelist([(x["in_node"], mapping[idx]) for idx, x in df.iterrows()] +\
                        [(x["out_node"], mapping[idx]) for idx, x in df.iterrows()], create_using=graph_type)
    
    nx.set_edge_attributes(G,{(x["in_node"], mapping[idx]):x["is_fraud"] for idx, x in df.iterrows()}, "label")
     
    nx.set_edge_attributes(G,{(x["out_node"], mapping[idx]):x["is_fraud"] for idx, x in df.iterrows()}, "label")
    
    nx.set_edge_attributes(G,{(x["in_node"], mapping[idx]):x["amt"] for idx, x in df.iterrows()}, "weight")
    
    nx.set_edge_attributes(G,{(x["out_node"], mapping[idx]):x["amt"] for idx, x in df.iterrows()}, "weight")
    
    
    return G
    

지도학습

0.3 / 0.2

df = throw(df, 0.3)

df_tr, df_tst = split_dataframe(df, 0.2)

df_tr.shape

(14014, 23)

df_tst.shape

(6006, 23)

df_tr.is_fraud.mean()

0.34287141429998574

df_tst.is_fraud.mean()

0.19996669996669997

df_, mask = concat(df_tr, df_tst)

G_down = build_graph_tripartite(df_)

G_down.number_of_edges(), G_down.number_of_nodes()

(40040, 21656)

edges = G_down.edges

sg.StellarGraph(edges=edges, edge_type_column="label")

from sklearn.model_selection import train_test_split


train_edges, test_edges, train_labels, test_labels = train_test_split(list(range(len(G_down.edges))), 
                                                                      list(nx.get_edge_attributes(G_down, "label").values()), 
                                                                      test_size=0.20, 
                                                                      random_state=42)

np.array(train_labels).mean(), np.array(test_labels).mean()

(0.2995442057942058, 0.30182317682317683)

edgs = list(G_down.edges)
train_graph = G_down.edge_subgraph([edgs[x] for x in train_edges]).copy()      
train_graph.add_nodes_from(list(set(G_down.nodes) - set(train_graph.nodes)))

import stellargraph as sg
from stellargraph.mapper import GraphSAGENodeGenerator
from stellargraph.layer import GraphSAGE
from sklearn.model_selection import train_test_split
from tensorflow import keras
from sklearn.preprocessing import LabelEncoder

graph = sg.StellarGraph(G_down)

/tmp/ipykernel_3324862/2906379029.py:1: DeprecationWarning: Constructing a StellarGraph directly from a NetworkX graph has been replaced by the `StellarGraph.from_networkx` function
  graph = sg.StellarGraph(G_down)

from node2vec import Node2Vec
from node2vec.edges import HadamardEmbedder, AverageEmbedder, WeightedL1Embedder, WeightedL2Embedder

node2vec_train = Node2Vec(train_graph, weight_key='weight')
model_train = node2vec_train.fit(window=10)

Generating walks (CPU: 1): 100%|██████████| 10/10 [00:44<00:00,  4.47s/it]

classes = [HadamardEmbedder, AverageEmbedder, WeightedL1Embedder, WeightedL2Embedder]
for cl in classes:
    embeddings_train = cl(keyed_vectors=model_train.wv) 

train_embeddings = [embeddings_train[str(edgs[x][0]), str(edgs[x][1])] for x in train_edges]
test_embeddings = [embeddings_train[str(edgs[x][0]), str(edgs[x][1])] for x in test_edges]

np.array(train_embeddings).shape

(32032, 128)

np.array(train_embeddings)

array([[8.79964884e-03, 1.64067835e-01, 7.30881765e-02, ...,
        1.81038718e-04, 1.96826290e-02, 6.05850630e-02],
       [3.18445265e-04, 9.41526145e-02, 2.41454929e-01, ...,
        5.79439476e-02, 9.22752500e-01, 2.49633682e-03],
       [1.09851332e-02, 4.12013801e-03, 1.61135435e-01, ...,
        1.24859456e-02, 2.32662242e-02, 1.07970215e-01],
       ...,
       [6.13867212e-03, 8.51532519e-02, 7.86146245e-07, ...,
        6.33678436e-02, 7.13208392e-02, 2.39914820e-01],
       [2.59715295e-03, 1.24797074e-03, 1.26128927e-01, ...,
        7.45704547e-02, 2.37582775e-04, 2.18050033e-01],
       [3.34992632e-02, 2.74142623e-03, 4.02400009e-02, ...,
        2.08475683e-02, 4.62760888e-02, 3.03567946e-01]], dtype=float32)

np.array(train_labels).shape

(32032,)

# DataFrame 생성
columns = [f'X_{i}' for i in range(np.array(train_embeddings).shape[1])]
df_data = pd.DataFrame(data=train_embeddings, columns=columns)

df_labels = pd.DataFrame(data=train_labels, columns=['label'])

# DataFrame 합치기
df = pd.concat([df_data, df_labels], axis=1)

df

	X_0	X_1	X_2	X_3	X_4	X_5	X_6	X_7	X_8	X_9	...	X_119	X_120	X_121	X_122	X_123	X_124	X_125	X_126	X_127	label
0	8.799649e-03	0.164068	7.308818e-02	0.007493	0.000196	0.037998	0.026417	0.030260	0.045050	0.047880	...	0.001140	0.082948	0.018060	0.041995	0.073492	0.011187	1.810387e-04	0.019683	0.060585	1
1	3.184453e-04	0.094153	2.414549e-01	0.016275	0.078201	0.015268	0.000326	0.331411	0.067376	0.061926	...	0.223335	0.000269	0.041893	0.051856	0.191570	0.000165	5.794395e-02	0.922752	0.002496	0
2	1.098513e-02	0.004120	1.611354e-01	0.548673	0.025099	0.001323	0.005411	0.005461	0.021607	0.094634	...	0.070138	0.038877	0.014588	0.003246	0.205133	0.064290	1.248595e-02	0.023266	0.107970	0
3	5.125533e-04	0.069008	2.394260e-02	0.038580	0.000345	0.058848	0.005284	0.168507	0.327026	0.002053	...	0.001018	0.037014	0.015680	0.193011	0.000812	0.001481	4.044086e-02	0.065931	0.131638	0
4	2.551593e-01	0.011703	4.802953e-02	0.020095	0.177149	0.022541	0.196618	0.041765	0.000156	0.158181	...	0.233692	0.104250	0.091681	0.002522	0.874027	0.003707	3.469664e-05	0.015489	0.005535	0
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
32027	1.361575e+00	0.023050	1.305244e-01	0.504093	0.014287	0.312939	0.000587	0.015945	0.049380	0.332058	...	0.002336	0.000683	0.191192	0.356860	0.227310	0.000661	1.307885e-01	0.023592	0.041412	0
32028	9.765987e-07	0.037045	1.420536e-02	0.026103	0.000017	0.116524	0.065952	0.054716	0.154811	0.011176	...	0.130505	0.006745	0.039672	0.052033	0.125356	0.015103	2.027648e-08	0.016023	0.169818	1
32029	6.138672e-03	0.085153	7.861462e-07	0.099862	0.408811	0.021244	0.006153	0.115799	0.000933	0.317840	...	0.177379	0.565970	0.022457	0.043790	0.059339	0.257367	6.336784e-02	0.071321	0.239915	0
32030	2.597153e-03	0.001248	1.261289e-01	0.000002	0.001143	0.248826	0.006462	0.016343	0.015120	0.108578	...	0.000021	0.000859	0.000894	0.496311	0.065650	0.007386	7.457045e-02	0.000238	0.218050	0
32031	3.349926e-02	0.002741	4.024000e-02	0.003076	0.000032	0.000168	0.065571	0.012371	0.112705	0.023244	...	0.138493	0.010161	0.032122	0.063049	0.028025	0.001975	2.084757e-02	0.046276	0.303568	1

32032 rows × 129 columns

label = np.array(train_labels)

predictr = TabularPredictor(label='label')

No path specified. Models will be saved in: "AutogluonModels/ag-20240124_110704/"

predictr.fit(df) 

Beginning AutoGluon training ...
AutoGluon will save models to "AutogluonModels/ag-20240124_110704/"
AutoGluon Version:  0.8.2
Python Version:     3.8.18
Operating System:   Linux
Platform Machine:   x86_64
Platform Version:   #38~22.04.1-Ubuntu SMP PREEMPT_DYNAMIC Thu Nov  2 18:01:13 UTC 2
Disk Space Avail:   623.47 GB / 982.82 GB (63.4%)
Train Data Rows:    32032
Train Data Columns: 128
Label Column: label
Preprocessing data ...
AutoGluon infers your prediction problem is: 'binary' (because only two unique label-values observed).
    2 unique label values:  [1, 0]
    If 'binary' is not the correct problem_type, please manually specify the problem_type parameter during predictor init (You may specify problem_type as one of: ['binary', 'multiclass', 'regression'])
Selected class <--> label mapping:  class 1 = 1, class 0 = 0
Using Feature Generators to preprocess the data ...
Fitting AutoMLPipelineFeatureGenerator...
    Available Memory:                    25264.96 MB
    Train Data (Original)  Memory Usage: 16.4 MB (0.1% of available memory)
    Inferring data type of each feature based on column values. Set feature_metadata_in to manually specify special dtypes of the features.
    Stage 1 Generators:
        Fitting AsTypeFeatureGenerator...
    Stage 2 Generators:
        Fitting FillNaFeatureGenerator...
    Stage 3 Generators:
        Fitting IdentityFeatureGenerator...
    Stage 4 Generators:
        Fitting DropUniqueFeatureGenerator...
    Stage 5 Generators:
        Fitting DropDuplicatesFeatureGenerator...
    Types of features in original data (raw dtype, special dtypes):
        ('float', []) : 128 | ['X_0', 'X_1', 'X_2', 'X_3', 'X_4', ...]
    Types of features in processed data (raw dtype, special dtypes):
        ('float', []) : 128 | ['X_0', 'X_1', 'X_2', 'X_3', 'X_4', ...]
    0.4s = Fit runtime
    128 features in original data used to generate 128 features in processed data.
    Train Data (Processed) Memory Usage: 16.4 MB (0.1% of available memory)
Data preprocessing and feature engineering runtime = 0.39s ...
AutoGluon will gauge predictive performance using evaluation metric: 'accuracy'
    To change this, specify the eval_metric parameter of Predictor()
Automatically generating train/validation split with holdout_frac=0.07804695304695304, Train Rows: 29532, Val Rows: 2500
User-specified model hyperparameters to be fit:
{
    'NN_TORCH': {},
    'GBM': [{'extra_trees': True, 'ag_args': {'name_suffix': 'XT'}}, {}, 'GBMLarge'],
    'CAT': {},
    'XGB': {},
    'FASTAI': {},
    'RF': [{'criterion': 'gini', 'ag_args': {'name_suffix': 'Gini', 'problem_types': ['binary', 'multiclass']}}, {'criterion': 'entropy', 'ag_args': {'name_suffix': 'Entr', 'problem_types': ['binary', 'multiclass']}}, {'criterion': 'squared_error', 'ag_args': {'name_suffix': 'MSE', 'problem_types': ['regression', 'quantile']}}],
    'XT': [{'criterion': 'gini', 'ag_args': {'name_suffix': 'Gini', 'problem_types': ['binary', 'multiclass']}}, {'criterion': 'entropy', 'ag_args': {'name_suffix': 'Entr', 'problem_types': ['binary', 'multiclass']}}, {'criterion': 'squared_error', 'ag_args': {'name_suffix': 'MSE', 'problem_types': ['regression', 'quantile']}}],
    'KNN': [{'weights': 'uniform', 'ag_args': {'name_suffix': 'Unif'}}, {'weights': 'distance', 'ag_args': {'name_suffix': 'Dist'}}],
}
Fitting 13 L1 models ...
Fitting model: KNeighborsUnif ...
Exception ignored on calling ctypes callback function: <function _ThreadpoolInfo._find_modules_with_dl_iterate_phdr.<locals>.match_module_callback at 0x7fcd7032fe50>
Traceback (most recent call last):
  File "/home/coco/anaconda3/envs/py38/lib/python3.8/site-packages/threadpoolctl.py", line 400, in match_module_callback
    self._make_module_from_path(filepath)
  File "/home/coco/anaconda3/envs/py38/lib/python3.8/site-packages/threadpoolctl.py", line 515, in _make_module_from_path
    module = module_class(filepath, prefix, user_api, internal_api)
  File "/home/coco/anaconda3/envs/py38/lib/python3.8/site-packages/threadpoolctl.py", line 606, in __init__
    self.version = self.get_version()
  File "/home/coco/anaconda3/envs/py38/lib/python3.8/site-packages/threadpoolctl.py", line 646, in get_version
    config = get_config().split()
AttributeError: 'NoneType' object has no attribute 'split'
    0.5904   = Validation score   (accuracy)
    0.07s    = Training   runtime
    0.18s    = Validation runtime
Fitting model: KNeighborsDist ...
Exception ignored on calling ctypes callback function: <function _ThreadpoolInfo._find_modules_with_dl_iterate_phdr.<locals>.match_module_callback at 0x7fcd7032fe50>
Traceback (most recent call last):
  File "/home/coco/anaconda3/envs/py38/lib/python3.8/site-packages/threadpoolctl.py", line 400, in match_module_callback
    self._make_module_from_path(filepath)
  File "/home/coco/anaconda3/envs/py38/lib/python3.8/site-packages/threadpoolctl.py", line 515, in _make_module_from_path
    module = module_class(filepath, prefix, user_api, internal_api)
  File "/home/coco/anaconda3/envs/py38/lib/python3.8/site-packages/threadpoolctl.py", line 606, in __init__
    self.version = self.get_version()
  File "/home/coco/anaconda3/envs/py38/lib/python3.8/site-packages/threadpoolctl.py", line 646, in get_version
    config = get_config().split()
AttributeError: 'NoneType' object has no attribute 'split'
    0.5904   = Validation score   (accuracy)
    0.07s    = Training   runtime
    0.08s    = Validation runtime
Fitting model: LightGBMXT ...
    0.726    = Validation score   (accuracy)
    1.07s    = Training   runtime
    0.01s    = Validation runtime
Fitting model: LightGBM ...
    0.7248   = Validation score   (accuracy)
    0.94s    = Training   runtime
    0.01s    = Validation runtime
Fitting model: RandomForestGini ...
    0.7236   = Validation score   (accuracy)
    6.19s    = Training   runtime
    0.04s    = Validation runtime
Fitting model: RandomForestEntr ...
    0.7248   = Validation score   (accuracy)
    8.16s    = Training   runtime
    0.04s    = Validation runtime
Fitting model: CatBoost ...
    0.7268   = Validation score   (accuracy)
    1.01s    = Training   runtime
    0.01s    = Validation runtime
Fitting model: ExtraTreesGini ...
    0.7288   = Validation score   (accuracy)
    1.02s    = Training   runtime
    0.05s    = Validation runtime
Fitting model: ExtraTreesEntr ...
    0.726    = Validation score   (accuracy)
    1.06s    = Training   runtime
    0.04s    = Validation runtime
Fitting model: NeuralNetFastAI ...
No improvement since epoch 5: early stopping
    0.742    = Validation score   (accuracy)
    14.83s   = Training   runtime
    0.03s    = Validation runtime
Fitting model: XGBoost ...
    0.7244   = Validation score   (accuracy)
    1.27s    = Training   runtime
    0.01s    = Validation runtime
Fitting model: NeuralNetTorch ...
    0.7432   = Validation score   (accuracy)
    11.94s   = Training   runtime
    0.11s    = Validation runtime
Fitting model: LightGBMLarge ...
    0.7268   = Validation score   (accuracy)
    1.99s    = Training   runtime
    0.01s    = Validation runtime
Fitting model: WeightedEnsemble_L2 ...
    0.7532   = Validation score   (accuracy)
    0.8s     = Training   runtime
    0.0s     = Validation runtime
AutoGluon training complete, total runtime = 52.58s ... Best model: "WeightedEnsemble_L2"
TabularPredictor saved. To load, use: predictor = TabularPredictor.load("AutogluonModels/ag-20240124_110704/")

<autogluon.tabular.predictor.predictor.TabularPredictor at 0x7fcd0b85c640>

test = np.array(test_embeddings)

test.shape

(8008, 128)

columns = [f'X_{i}' for i in range(test.shape[1])]

# DataFrame 생성
test_df = pd.DataFrame(data=test, columns=columns)

# DataFrame 확인
print(test_df.head())

        X_0       X_1       X_2       X_3       X_4       X_5       X_6  \
0  0.413663  2.075492  0.097850  0.431519  0.026412  0.046006  0.139835   
1  0.012225  0.000547  0.024713  0.703247  0.020913  0.419119  0.352671   
2  0.195774  0.000752  0.009002  0.204248  0.070720  0.906959  0.507191   
3  0.260361  0.329479  0.229454  0.023667  0.001113  0.002806  0.034591   
4  0.427026  0.024197  0.695513  0.057896  0.270117  0.026265  0.320806   

        X_7       X_8       X_9  ...     X_118     X_119     X_120     X_121  \
0  0.014063  0.473687  0.014283  ...  0.048837  0.010120  0.005523  0.961210   
1  1.669539  0.126232  0.027797  ...  0.728475  0.378981  1.377197  0.203648   
2  0.267592  0.115088  0.409537  ...  0.295417  0.440117  0.088578  0.060138   
3  0.045546  0.623125  0.282554  ...  0.174775  0.159482  0.001048  0.035379   
4  0.029051  0.263759  0.028128  ...  0.008239  0.063498  0.810952  0.013463   

      X_122     X_123     X_124     X_125     X_126     X_127  
0  0.026908  0.000516  0.054398  0.188361  0.007897  0.495728  
1  0.086590  0.230660  0.000614  1.211448  0.513626  0.000002  
2  0.004575  1.479674  0.131626  0.928811  0.315709  0.000010  
3  0.640979  0.329148  0.003226  0.278314  0.002369  0.150809  
4  0.295149  0.006146  0.608962  0.076195  0.106696  0.122096  

[5 rows x 128 columns]

predictr.predict(test_df).mean()

Exception ignored on calling ctypes callback function: <function _ThreadpoolInfo._find_modules_with_dl_iterate_phdr.<locals>.match_module_callback at 0x7fcd8f12cb80>
Traceback (most recent call last):
  File "/home/coco/anaconda3/envs/py38/lib/python3.8/site-packages/threadpoolctl.py", line 400, in match_module_callback
    self._make_module_from_path(filepath)
  File "/home/coco/anaconda3/envs/py38/lib/python3.8/site-packages/threadpoolctl.py", line 515, in _make_module_from_path
    module = module_class(filepath, prefix, user_api, internal_api)
  File "/home/coco/anaconda3/envs/py38/lib/python3.8/site-packages/threadpoolctl.py", line 606, in __init__
    self.version = self.get_version()
  File "/home/coco/anaconda3/envs/py38/lib/python3.8/site-packages/threadpoolctl.py", line 646, in get_version
    config = get_config().split()
AttributeError: 'NoneType' object has no attribute 'split'

0.0241008991008991

y = np.array(test_labels)

yhat = predictr.predict(test_df)

Exception ignored on calling ctypes callback function: <function _ThreadpoolInfo._find_modules_with_dl_iterate_phdr.<locals>.match_module_callback at 0x7fcd0b8c03a0>
Traceback (most recent call last):
  File "/home/coco/anaconda3/envs/py38/lib/python3.8/site-packages/threadpoolctl.py", line 400, in match_module_callback
    self._make_module_from_path(filepath)
  File "/home/coco/anaconda3/envs/py38/lib/python3.8/site-packages/threadpoolctl.py", line 515, in _make_module_from_path
    module = module_class(filepath, prefix, user_api, internal_api)
  File "/home/coco/anaconda3/envs/py38/lib/python3.8/site-packages/threadpoolctl.py", line 606, in __init__
    self.version = self.get_version()
  File "/home/coco/anaconda3/envs/py38/lib/python3.8/site-packages/threadpoolctl.py", line 646, in get_version
    config = get_config().split()
AttributeError: 'NoneType' object has no attribute 'split'

# sklearn
import sklearn
from sklearn import metrics
from sklearn.metrics import precision_score, recall_score, f1_score

    

evaluation(y,yhat)

	accuracy_score	precision_score	recall_score	f1_score	roc_auc_score
0	0.690684	0.264249	0.021375	0.03955	0.498058

---

0.3 / 0.3

df = throw(df, 0.3)

df_tr, df_tst = split_dataframe(df, 0.3)

df_, mask = concat(df_tr, df_tst)

G_down = build_graph_tripartite(df_)

train_edges, test_edges, train_labels, test_labels = train_test_split(list(range(len(G_down.edges))), 
                                                                      list(nx.get_edge_attributes(G_down, "label").values()), 
                                                                      test_size=0.20, 
                                                                      random_state=42)

edgs = list(G_down.edges)
train_graph = G_down.edge_subgraph([edgs[x] for x in train_edges]).copy()      
train_graph.add_nodes_from(list(set(G_down.nodes) - set(train_graph.nodes)))



node2vec_train = Node2Vec(train_graph, weight_key='weight')
model_train = node2vec_train.fit(window=10)

classes = [HadamardEmbedder, AverageEmbedder, WeightedL1Embedder, WeightedL2Embedder]
for cl in classes:
    embeddings_train = cl(keyed_vectors=model_train.wv) 

train_embeddings = [embeddings_train[str(edgs[x][0]), str(edgs[x][1])] for x in train_edges]
test_embeddings = [embeddings_train[str(edgs[x][0]), str(edgs[x][1])] for x in test_edges]


# DataFrame 생성
columns = [f'X_{i}' for i in range(np.array(train_embeddings).shape[1])]
df_data = pd.DataFrame(data=train_embeddings, columns=columns)

df_labels = pd.DataFrame(data=train_labels, columns=['label'])

# DataFrame 합치기
df = pd.concat([df_data, df_labels], axis=1)


label = np.array(train_labels)

Generating walks (CPU: 1): 100%|██████████| 10/10 [00:46<00:00,  4.61s/it]

predictr = TabularPredictor(label='label')

No path specified. Models will be saved in: "AutogluonModels/ag-20240124_112803/"

predictr.fit(df) 

Beginning AutoGluon training ...
AutoGluon will save models to "AutogluonModels/ag-20240124_112803/"
AutoGluon Version:  0.8.2
Python Version:     3.8.18
Operating System:   Linux
Platform Machine:   x86_64
Platform Version:   #38~22.04.1-Ubuntu SMP PREEMPT_DYNAMIC Thu Nov  2 18:01:13 UTC 2
Disk Space Avail:   622.65 GB / 982.82 GB (63.4%)
Train Data Rows:    32032
Train Data Columns: 128
Label Column: label
Preprocessing data ...
AutoGluon infers your prediction problem is: 'binary' (because only two unique label-values observed).
    2 unique label values:  [0, 1]
    If 'binary' is not the correct problem_type, please manually specify the problem_type parameter during predictor init (You may specify problem_type as one of: ['binary', 'multiclass', 'regression'])
Selected class <--> label mapping:  class 1 = 1, class 0 = 0
Using Feature Generators to preprocess the data ...
Fitting AutoMLPipelineFeatureGenerator...
    Available Memory:                    24167.01 MB
    Train Data (Original)  Memory Usage: 16.4 MB (0.1% of available memory)
    Inferring data type of each feature based on column values. Set feature_metadata_in to manually specify special dtypes of the features.
    Stage 1 Generators:
        Fitting AsTypeFeatureGenerator...
    Stage 2 Generators:
        Fitting FillNaFeatureGenerator...
    Stage 3 Generators:
        Fitting IdentityFeatureGenerator...
    Stage 4 Generators:
        Fitting DropUniqueFeatureGenerator...
    Stage 5 Generators:
        Fitting DropDuplicatesFeatureGenerator...
    Types of features in original data (raw dtype, special dtypes):
        ('float', []) : 128 | ['X_0', 'X_1', 'X_2', 'X_3', 'X_4', ...]
    Types of features in processed data (raw dtype, special dtypes):
        ('float', []) : 128 | ['X_0', 'X_1', 'X_2', 'X_3', 'X_4', ...]
    0.4s = Fit runtime
    128 features in original data used to generate 128 features in processed data.
    Train Data (Processed) Memory Usage: 16.4 MB (0.1% of available memory)
Data preprocessing and feature engineering runtime = 0.38s ...
AutoGluon will gauge predictive performance using evaluation metric: 'accuracy'
    To change this, specify the eval_metric parameter of Predictor()
Automatically generating train/validation split with holdout_frac=0.07804695304695304, Train Rows: 29532, Val Rows: 2500
User-specified model hyperparameters to be fit:
{
    'NN_TORCH': {},
    'GBM': [{'extra_trees': True, 'ag_args': {'name_suffix': 'XT'}}, {}, 'GBMLarge'],
    'CAT': {},
    'XGB': {},
    'FASTAI': {},
    'RF': [{'criterion': 'gini', 'ag_args': {'name_suffix': 'Gini', 'problem_types': ['binary', 'multiclass']}}, {'criterion': 'entropy', 'ag_args': {'name_suffix': 'Entr', 'problem_types': ['binary', 'multiclass']}}, {'criterion': 'squared_error', 'ag_args': {'name_suffix': 'MSE', 'problem_types': ['regression', 'quantile']}}],
    'XT': [{'criterion': 'gini', 'ag_args': {'name_suffix': 'Gini', 'problem_types': ['binary', 'multiclass']}}, {'criterion': 'entropy', 'ag_args': {'name_suffix': 'Entr', 'problem_types': ['binary', 'multiclass']}}, {'criterion': 'squared_error', 'ag_args': {'name_suffix': 'MSE', 'problem_types': ['regression', 'quantile']}}],
    'KNN': [{'weights': 'uniform', 'ag_args': {'name_suffix': 'Unif'}}, {'weights': 'distance', 'ag_args': {'name_suffix': 'Dist'}}],
}
Fitting 13 L1 models ...
Fitting model: KNeighborsUnif ...
Exception ignored on calling ctypes callback function: <function _ThreadpoolInfo._find_modules_with_dl_iterate_phdr.<locals>.match_module_callback at 0x7fcd0b8c00d0>
Traceback (most recent call last):
  File "/home/coco/anaconda3/envs/py38/lib/python3.8/site-packages/threadpoolctl.py", line 400, in match_module_callback
    self._make_module_from_path(filepath)
  File "/home/coco/anaconda3/envs/py38/lib/python3.8/site-packages/threadpoolctl.py", line 515, in _make_module_from_path
    module = module_class(filepath, prefix, user_api, internal_api)
  File "/home/coco/anaconda3/envs/py38/lib/python3.8/site-packages/threadpoolctl.py", line 606, in __init__
    self.version = self.get_version()
  File "/home/coco/anaconda3/envs/py38/lib/python3.8/site-packages/threadpoolctl.py", line 646, in get_version
    config = get_config().split()
AttributeError: 'NoneType' object has no attribute 'split'
    0.594    = Validation score   (accuracy)
    0.07s    = Training   runtime
    0.46s    = Validation runtime
Fitting model: KNeighborsDist ...
Exception ignored on calling ctypes callback function: <function _ThreadpoolInfo._find_modules_with_dl_iterate_phdr.<locals>.match_module_callback at 0x7fcd0b8c00d0>
Traceback (most recent call last):
  File "/home/coco/anaconda3/envs/py38/lib/python3.8/site-packages/threadpoolctl.py", line 400, in match_module_callback
    self._make_module_from_path(filepath)
  File "/home/coco/anaconda3/envs/py38/lib/python3.8/site-packages/threadpoolctl.py", line 515, in _make_module_from_path
    module = module_class(filepath, prefix, user_api, internal_api)
  File "/home/coco/anaconda3/envs/py38/lib/python3.8/site-packages/threadpoolctl.py", line 606, in __init__
    self.version = self.get_version()
  File "/home/coco/anaconda3/envs/py38/lib/python3.8/site-packages/threadpoolctl.py", line 646, in get_version
    config = get_config().split()
AttributeError: 'NoneType' object has no attribute 'split'
    0.5956   = Validation score   (accuracy)
    0.09s    = Training   runtime
    0.49s    = Validation runtime
Fitting model: LightGBMXT ...
    0.7392   = Validation score   (accuracy)
    1.03s    = Training   runtime
    0.0s     = Validation runtime
Fitting model: LightGBM ...
    0.74     = Validation score   (accuracy)
    1.08s    = Training   runtime
    0.01s    = Validation runtime
Fitting model: RandomForestGini ...
    0.7304   = Validation score   (accuracy)
    6.74s    = Training   runtime
    0.05s    = Validation runtime
Fitting model: RandomForestEntr ...
    0.7268   = Validation score   (accuracy)
    8.77s    = Training   runtime
    0.05s    = Validation runtime
Fitting model: CatBoost ...
    0.7396   = Validation score   (accuracy)
    1.06s    = Training   runtime
    0.01s    = Validation runtime
Fitting model: ExtraTreesGini ...
    0.7328   = Validation score   (accuracy)
    1.12s    = Training   runtime
    0.05s    = Validation runtime
Fitting model: ExtraTreesEntr ...
    0.7312   = Validation score   (accuracy)
    1.18s    = Training   runtime
    0.05s    = Validation runtime
Fitting model: NeuralNetFastAI ...
No improvement since epoch 4: early stopping
    0.758    = Validation score   (accuracy)
    14.33s   = Training   runtime
    0.02s    = Validation runtime
Fitting model: XGBoost ...
    0.7376   = Validation score   (accuracy)
    1.16s    = Training   runtime
    0.01s    = Validation runtime
Fitting model: NeuralNetTorch ...
    0.7464   = Validation score   (accuracy)
    10.13s   = Training   runtime
    0.09s    = Validation runtime
Fitting model: LightGBMLarge ...
    0.738    = Validation score   (accuracy)
    2.1s     = Training   runtime
    0.01s    = Validation runtime
Fitting model: WeightedEnsemble_L2 ...
    0.7708   = Validation score   (accuracy)
    0.8s     = Training   runtime
    0.0s     = Validation runtime
AutoGluon training complete, total runtime = 52.42s ... Best model: "WeightedEnsemble_L2"
TabularPredictor saved. To load, use: predictor = TabularPredictor.load("AutogluonModels/ag-20240124_112803/")

<autogluon.tabular.predictor.predictor.TabularPredictor at 0x7fcd8f0e0190>

test = np.array(test_embeddings)

columns = [f'X_{i}' for i in range(test.shape[1])]

# DataFrame 생성
test_df = pd.DataFrame(data=test, columns=columns)

y = np.array(test_labels)

yhat = predictr.predict(test_df)

Exception ignored on calling ctypes callback function: <function _ThreadpoolInfo._find_modules_with_dl_iterate_phdr.<locals>.match_module_callback at 0x7fccf01554c0>
Traceback (most recent call last):
  File "/home/coco/anaconda3/envs/py38/lib/python3.8/site-packages/threadpoolctl.py", line 400, in match_module_callback
    self._make_module_from_path(filepath)
  File "/home/coco/anaconda3/envs/py38/lib/python3.8/site-packages/threadpoolctl.py", line 515, in _make_module_from_path
    module = module_class(filepath, prefix, user_api, internal_api)
  File "/home/coco/anaconda3/envs/py38/lib/python3.8/site-packages/threadpoolctl.py", line 606, in __init__
    self.version = self.get_version()
  File "/home/coco/anaconda3/envs/py38/lib/python3.8/site-packages/threadpoolctl.py", line 646, in get_version
    config = get_config().split()
AttributeError: 'NoneType' object has no attribute 'split'

evaluation(y,yhat)

/home/coco/anaconda3/envs/py38/lib/python3.8/site-packages/sklearn/metrics/_classification.py:1344: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.
  _warn_prf(average, modifier, msg_start, len(result))

	accuracy_score	precision_score	recall_score	f1_score	roc_auc_score
0	0.69493	0.0	0.0	0.0	0.5

---

0.3 / 0.4

df = pd.read_csv("~/Desktop/fraudTrain.csv")

df = throw(df, 0.3)
df_tr, df_tst = split_dataframe(df, 0.4)
df_, mask = concat(df_tr, df_tst)
G_down = build_graph_tripartite(df_)
train_edges, test_edges, train_labels, test_labels = train_test_split(list(range(len(G_down.edges))), 
                                                                      list(nx.get_edge_attributes(G_down, "label").values()), 
                                                                      test_size=0.20, 
                                                                      random_state=42)
edgs = list(G_down.edges)
train_graph = G_down.edge_subgraph([edgs[x] for x in train_edges]).copy()      
train_graph.add_nodes_from(list(set(G_down.nodes) - set(train_graph.nodes)))


node2vec_train = Node2Vec(train_graph, weight_key='weight')
model_train = node2vec_train.fit(window=10)
classes = [HadamardEmbedder, AverageEmbedder, WeightedL1Embedder, WeightedL2Embedder]
for cl in classes:
    embeddings_train = cl(keyed_vectors=model_train.wv) 

train_embeddings = [embeddings_train[str(edgs[x][0]), str(edgs[x][1])] for x in train_edges]
test_embeddings = [embeddings_train[str(edgs[x][0]), str(edgs[x][1])] for x in test_edges]


# DataFrame 생성
columns = [f'X_{i}' for i in range(np.array(train_embeddings).shape[1])]
df_data = pd.DataFrame(data=train_embeddings, columns=columns)
df_labels = pd.DataFrame(data=train_labels, columns=['label'])

df = pd.concat([df_data, df_labels], axis=1)
label = np.array(train_labels)

predictr = TabularPredictor(label='label')

predictr.fit(df) 

test = np.array(test_embeddings)

columns = [f'X_{i}' for i in range(test.shape[1])]

# DataFrame 생성
test_df = pd.DataFrame(data=test, columns=columns)

y = np.array(test_labels)
yhat = predictr.predict(test_df)

evaluation(y,yhat)

Generating walks (CPU: 1): 100%|██████████| 10/10 [00:48<00:00,  4.86s/it]
No path specified. Models will be saved in: "AutogluonModels/ag-20240124_113911/"
Beginning AutoGluon training ...
AutoGluon will save models to "AutogluonModels/ag-20240124_113911/"
AutoGluon Version:  0.8.2
Python Version:     3.8.18
Operating System:   Linux
Platform Machine:   x86_64
Platform Version:   #38~22.04.1-Ubuntu SMP PREEMPT_DYNAMIC Thu Nov  2 18:01:13 UTC 2
Disk Space Avail:   621.83 GB / 982.82 GB (63.3%)
Train Data Rows:    32032
Train Data Columns: 128
Label Column: label
Preprocessing data ...
AutoGluon infers your prediction problem is: 'binary' (because only two unique label-values observed).
    2 unique label values:  [0, 1]
    If 'binary' is not the correct problem_type, please manually specify the problem_type parameter during predictor init (You may specify problem_type as one of: ['binary', 'multiclass', 'regression'])
Selected class <--> label mapping:  class 1 = 1, class 0 = 0
Using Feature Generators to preprocess the data ...
Fitting AutoMLPipelineFeatureGenerator...
    Available Memory:                    23867.75 MB
    Train Data (Original)  Memory Usage: 16.4 MB (0.1% of available memory)
    Inferring data type of each feature based on column values. Set feature_metadata_in to manually specify special dtypes of the features.
    Stage 1 Generators:
        Fitting AsTypeFeatureGenerator...
    Stage 2 Generators:
        Fitting FillNaFeatureGenerator...
    Stage 3 Generators:
        Fitting IdentityFeatureGenerator...
    Stage 4 Generators:
        Fitting DropUniqueFeatureGenerator...
    Stage 5 Generators:
        Fitting DropDuplicatesFeatureGenerator...
    Types of features in original data (raw dtype, special dtypes):
        ('float', []) : 128 | ['X_0', 'X_1', 'X_2', 'X_3', 'X_4', ...]
    Types of features in processed data (raw dtype, special dtypes):
        ('float', []) : 128 | ['X_0', 'X_1', 'X_2', 'X_3', 'X_4', ...]
    0.3s = Fit runtime
    128 features in original data used to generate 128 features in processed data.
    Train Data (Processed) Memory Usage: 16.4 MB (0.1% of available memory)
Data preprocessing and feature engineering runtime = 0.36s ...
AutoGluon will gauge predictive performance using evaluation metric: 'accuracy'
    To change this, specify the eval_metric parameter of Predictor()
Automatically generating train/validation split with holdout_frac=0.07804695304695304, Train Rows: 29532, Val Rows: 2500
User-specified model hyperparameters to be fit:
{
    'NN_TORCH': {},
    'GBM': [{'extra_trees': True, 'ag_args': {'name_suffix': 'XT'}}, {}, 'GBMLarge'],
    'CAT': {},
    'XGB': {},
    'FASTAI': {},
    'RF': [{'criterion': 'gini', 'ag_args': {'name_suffix': 'Gini', 'problem_types': ['binary', 'multiclass']}}, {'criterion': 'entropy', 'ag_args': {'name_suffix': 'Entr', 'problem_types': ['binary', 'multiclass']}}, {'criterion': 'squared_error', 'ag_args': {'name_suffix': 'MSE', 'problem_types': ['regression', 'quantile']}}],
    'XT': [{'criterion': 'gini', 'ag_args': {'name_suffix': 'Gini', 'problem_types': ['binary', 'multiclass']}}, {'criterion': 'entropy', 'ag_args': {'name_suffix': 'Entr', 'problem_types': ['binary', 'multiclass']}}, {'criterion': 'squared_error', 'ag_args': {'name_suffix': 'MSE', 'problem_types': ['regression', 'quantile']}}],
    'KNN': [{'weights': 'uniform', 'ag_args': {'name_suffix': 'Unif'}}, {'weights': 'distance', 'ag_args': {'name_suffix': 'Dist'}}],
}
Fitting 13 L1 models ...
Fitting model: KNeighborsUnif ...
Exception ignored on calling ctypes callback function: <function _ThreadpoolInfo._find_modules_with_dl_iterate_phdr.<locals>.match_module_callback at 0x7fcd0b85e940>
Traceback (most recent call last):
  File "/home/coco/anaconda3/envs/py38/lib/python3.8/site-packages/threadpoolctl.py", line 400, in match_module_callback
    self._make_module_from_path(filepath)
  File "/home/coco/anaconda3/envs/py38/lib/python3.8/site-packages/threadpoolctl.py", line 515, in _make_module_from_path
    module = module_class(filepath, prefix, user_api, internal_api)
  File "/home/coco/anaconda3/envs/py38/lib/python3.8/site-packages/threadpoolctl.py", line 606, in __init__
    self.version = self.get_version()
  File "/home/coco/anaconda3/envs/py38/lib/python3.8/site-packages/threadpoolctl.py", line 646, in get_version
    config = get_config().split()
AttributeError: 'NoneType' object has no attribute 'split'
    0.5988   = Validation score   (accuracy)
    0.07s    = Training   runtime
    0.4s     = Validation runtime
Fitting model: KNeighborsDist ...
Exception ignored on calling ctypes callback function: <function _ThreadpoolInfo._find_modules_with_dl_iterate_phdr.<locals>.match_module_callback at 0x7fcd0b85e8b0>
Traceback (most recent call last):
  File "/home/coco/anaconda3/envs/py38/lib/python3.8/site-packages/threadpoolctl.py", line 400, in match_module_callback
    self._make_module_from_path(filepath)
  File "/home/coco/anaconda3/envs/py38/lib/python3.8/site-packages/threadpoolctl.py", line 515, in _make_module_from_path
    module = module_class(filepath, prefix, user_api, internal_api)
  File "/home/coco/anaconda3/envs/py38/lib/python3.8/site-packages/threadpoolctl.py", line 606, in __init__
    self.version = self.get_version()
  File "/home/coco/anaconda3/envs/py38/lib/python3.8/site-packages/threadpoolctl.py", line 646, in get_version
    config = get_config().split()
AttributeError: 'NoneType' object has no attribute 'split'
    0.5992   = Validation score   (accuracy)
    0.08s    = Training   runtime
    0.47s    = Validation runtime
Fitting model: LightGBMXT ...
    0.738    = Validation score   (accuracy)
    0.99s    = Training   runtime
    0.0s     = Validation runtime
Fitting model: LightGBM ...
    0.738    = Validation score   (accuracy)
    1.0s     = Training   runtime
    0.01s    = Validation runtime
Fitting model: RandomForestGini ...
    0.7324   = Validation score   (accuracy)
    6.23s    = Training   runtime
    0.04s    = Validation runtime
Fitting model: RandomForestEntr ...
    0.7292   = Validation score   (accuracy)
    8.22s    = Training   runtime
    0.04s    = Validation runtime
Fitting model: CatBoost ...
    0.7364   = Validation score   (accuracy)
    1.0s     = Training   runtime
    0.01s    = Validation runtime
Fitting model: ExtraTreesGini ...
    0.7372   = Validation score   (accuracy)
    1.07s    = Training   runtime
    0.05s    = Validation runtime
Fitting model: ExtraTreesEntr ...
    0.7324   = Validation score   (accuracy)
    1.11s    = Training   runtime
    0.05s    = Validation runtime
Fitting model: NeuralNetFastAI ...
No improvement since epoch 8: early stopping
    0.756    = Validation score   (accuracy)
    15.57s   = Training   runtime
    0.02s    = Validation runtime
Fitting model: XGBoost ...
    0.7352   = Validation score   (accuracy)
    1.1s     = Training   runtime
    0.01s    = Validation runtime
Fitting model: NeuralNetTorch ...
    0.7552   = Validation score   (accuracy)
    10.67s   = Training   runtime
    0.09s    = Validation runtime
Fitting model: LightGBMLarge ...
    0.738    = Validation score   (accuracy)
    2.21s    = Training   runtime
    0.01s    = Validation runtime
Fitting model: WeightedEnsemble_L2 ...
    0.7688   = Validation score   (accuracy)
    0.84s    = Training   runtime
    0.0s     = Validation runtime
AutoGluon training complete, total runtime = 52.62s ... Best model: "WeightedEnsemble_L2"
TabularPredictor saved. To load, use: predictor = TabularPredictor.load("AutogluonModels/ag-20240124_113911/")
Exception ignored on calling ctypes callback function: <function _ThreadpoolInfo._find_modules_with_dl_iterate_phdr.<locals>.match_module_callback at 0x7fcd0b85e790>
Traceback (most recent call last):
  File "/home/coco/anaconda3/envs/py38/lib/python3.8/site-packages/threadpoolctl.py", line 400, in match_module_callback
    self._make_module_from_path(filepath)
  File "/home/coco/anaconda3/envs/py38/lib/python3.8/site-packages/threadpoolctl.py", line 515, in _make_module_from_path
    module = module_class(filepath, prefix, user_api, internal_api)
  File "/home/coco/anaconda3/envs/py38/lib/python3.8/site-packages/threadpoolctl.py", line 606, in __init__
    self.version = self.get_version()
  File "/home/coco/anaconda3/envs/py38/lib/python3.8/site-packages/threadpoolctl.py", line 646, in get_version
    config = get_config().split()
AttributeError: 'NoneType' object has no attribute 'split'

	accuracy_score	precision_score	recall_score	f1_score	roc_auc_score
0	0.688686	0.350318	0.045852	0.081091	0.504741

---

0.4 / 0.4

df = pd.read_csv("~/Desktop/fraudTrain.csv")

df = throw(df, 0.4)
df_tr, df_tst = split_dataframe(df, 0.4)
df_, mask = concat(df_tr, df_tst)
G_down = build_graph_tripartite(df_)
train_edges, test_edges, train_labels, test_labels = train_test_split(list(range(len(G_down.edges))), 
                                                                      list(nx.get_edge_attributes(G_down, "label").values()), 
                                                                      test_size=0.20, 
                                                                      random_state=42)
edgs = list(G_down.edges)
train_graph = G_down.edge_subgraph([edgs[x] for x in train_edges]).copy()      
train_graph.add_nodes_from(list(set(G_down.nodes) - set(train_graph.nodes)))


node2vec_train = Node2Vec(train_graph, weight_key='weight')
model_train = node2vec_train.fit(window=10)
classes = [HadamardEmbedder, AverageEmbedder, WeightedL1Embedder, WeightedL2Embedder]
for cl in classes:
    embeddings_train = cl(keyed_vectors=model_train.wv) 

train_embeddings = [embeddings_train[str(edgs[x][0]), str(edgs[x][1])] for x in train_edges]
test_embeddings = [embeddings_train[str(edgs[x][0]), str(edgs[x][1])] for x in test_edges]


# DataFrame 생성
columns = [f'X_{i}' for i in range(np.array(train_embeddings).shape[1])]
df_data = pd.DataFrame(data=train_embeddings, columns=columns)
df_labels = pd.DataFrame(data=train_labels, columns=['label'])

df = pd.concat([df_data, df_labels], axis=1)
label = np.array(train_labels)

predictr = TabularPredictor(label='label')

predictr.fit(df) 

test = np.array(test_embeddings)

columns = [f'X_{i}' for i in range(test.shape[1])]

# DataFrame 생성
test_df = pd.DataFrame(data=test, columns=columns)

y = np.array(test_labels)
yhat = predictr.predict(test_df)

evaluation(y,yhat)

Generating walks (CPU: 1): 100%|██████████| 10/10 [00:32<00:00,  3.23s/it]
No path specified. Models will be saved in: "AutogluonModels/ag-20240125_002023/"
Beginning AutoGluon training ...
AutoGluon will save models to "AutogluonModels/ag-20240125_002023/"
AutoGluon Version:  0.8.2
Python Version:     3.8.18
Operating System:   Linux
Platform Machine:   x86_64
Platform Version:   #38~22.04.1-Ubuntu SMP PREEMPT_DYNAMIC Thu Nov  2 18:01:13 UTC 2
Disk Space Avail:   620.00 GB / 982.82 GB (63.1%)
Train Data Rows:    24024
Train Data Columns: 128
Label Column: label
Preprocessing data ...
AutoGluon infers your prediction problem is: 'binary' (because only two unique label-values observed).
    2 unique label values:  [1, 0]
    If 'binary' is not the correct problem_type, please manually specify the problem_type parameter during predictor init (You may specify problem_type as one of: ['binary', 'multiclass', 'regression'])
Selected class <--> label mapping:  class 1 = 1, class 0 = 0
Using Feature Generators to preprocess the data ...
Fitting AutoMLPipelineFeatureGenerator...
    Available Memory:                    24334.27 MB
    Train Data (Original)  Memory Usage: 12.3 MB (0.1% of available memory)
    Inferring data type of each feature based on column values. Set feature_metadata_in to manually specify special dtypes of the features.
    Stage 1 Generators:
        Fitting AsTypeFeatureGenerator...
    Stage 2 Generators:
        Fitting FillNaFeatureGenerator...
    Stage 3 Generators:
        Fitting IdentityFeatureGenerator...
    Stage 4 Generators:
        Fitting DropUniqueFeatureGenerator...
    Stage 5 Generators:
        Fitting DropDuplicatesFeatureGenerator...
    Types of features in original data (raw dtype, special dtypes):
        ('float', []) : 128 | ['X_0', 'X_1', 'X_2', 'X_3', 'X_4', ...]
    Types of features in processed data (raw dtype, special dtypes):
        ('float', []) : 128 | ['X_0', 'X_1', 'X_2', 'X_3', 'X_4', ...]
    0.7s = Fit runtime
    128 features in original data used to generate 128 features in processed data.
    Train Data (Processed) Memory Usage: 12.3 MB (0.1% of available memory)
Data preprocessing and feature engineering runtime = 0.7s ...
AutoGluon will gauge predictive performance using evaluation metric: 'accuracy'
    To change this, specify the eval_metric parameter of Predictor()
Automatically generating train/validation split with holdout_frac=0.1, Train Rows: 21621, Val Rows: 2403
User-specified model hyperparameters to be fit:
{
    'NN_TORCH': {},
    'GBM': [{'extra_trees': True, 'ag_args': {'name_suffix': 'XT'}}, {}, 'GBMLarge'],
    'CAT': {},
    'XGB': {},
    'FASTAI': {},
    'RF': [{'criterion': 'gini', 'ag_args': {'name_suffix': 'Gini', 'problem_types': ['binary', 'multiclass']}}, {'criterion': 'entropy', 'ag_args': {'name_suffix': 'Entr', 'problem_types': ['binary', 'multiclass']}}, {'criterion': 'squared_error', 'ag_args': {'name_suffix': 'MSE', 'problem_types': ['regression', 'quantile']}}],
    'XT': [{'criterion': 'gini', 'ag_args': {'name_suffix': 'Gini', 'problem_types': ['binary', 'multiclass']}}, {'criterion': 'entropy', 'ag_args': {'name_suffix': 'Entr', 'problem_types': ['binary', 'multiclass']}}, {'criterion': 'squared_error', 'ag_args': {'name_suffix': 'MSE', 'problem_types': ['regression', 'quantile']}}],
    'KNN': [{'weights': 'uniform', 'ag_args': {'name_suffix': 'Unif'}}, {'weights': 'distance', 'ag_args': {'name_suffix': 'Dist'}}],
}
Fitting 13 L1 models ...
Fitting model: KNeighborsUnif ...
Exception ignored on calling ctypes callback function: <function _ThreadpoolInfo._find_modules_with_dl_iterate_phdr.<locals>.match_module_callback at 0x7fcd0b85e1f0>
Traceback (most recent call last):
  File "/home/coco/anaconda3/envs/py38/lib/python3.8/site-packages/threadpoolctl.py", line 400, in match_module_callback
    self._make_module_from_path(filepath)
  File "/home/coco/anaconda3/envs/py38/lib/python3.8/site-packages/threadpoolctl.py", line 515, in _make_module_from_path
    module = module_class(filepath, prefix, user_api, internal_api)
  File "/home/coco/anaconda3/envs/py38/lib/python3.8/site-packages/threadpoolctl.py", line 606, in __init__
    self.version = self.get_version()
  File "/home/coco/anaconda3/envs/py38/lib/python3.8/site-packages/threadpoolctl.py", line 646, in get_version
    config = get_config().split()
AttributeError: 'NoneType' object has no attribute 'split'
    0.5568   = Validation score   (accuracy)
    0.06s    = Training   runtime
    0.38s    = Validation runtime
Fitting model: KNeighborsDist ...
Exception ignored on calling ctypes callback function: <function _ThreadpoolInfo._find_modules_with_dl_iterate_phdr.<locals>.match_module_callback at 0x7fccf82ed550>
Traceback (most recent call last):
  File "/home/coco/anaconda3/envs/py38/lib/python3.8/site-packages/threadpoolctl.py", line 400, in match_module_callback
    self._make_module_from_path(filepath)
  File "/home/coco/anaconda3/envs/py38/lib/python3.8/site-packages/threadpoolctl.py", line 515, in _make_module_from_path
    module = module_class(filepath, prefix, user_api, internal_api)
  File "/home/coco/anaconda3/envs/py38/lib/python3.8/site-packages/threadpoolctl.py", line 606, in __init__
    self.version = self.get_version()
  File "/home/coco/anaconda3/envs/py38/lib/python3.8/site-packages/threadpoolctl.py", line 646, in get_version
    config = get_config().split()
AttributeError: 'NoneType' object has no attribute 'split'
    0.5576   = Validation score   (accuracy)
    0.07s    = Training   runtime
    0.45s    = Validation runtime
Fitting model: LightGBMXT ...
    0.6746   = Validation score   (accuracy)
    1.38s    = Training   runtime
    0.01s    = Validation runtime
Fitting model: LightGBM ...
    0.6767   = Validation score   (accuracy)
    1.73s    = Training   runtime
    0.01s    = Validation runtime
Fitting model: RandomForestGini ...
    0.6479   = Validation score   (accuracy)
    4.39s    = Training   runtime
    0.04s    = Validation runtime
Fitting model: RandomForestEntr ...
    0.6583   = Validation score   (accuracy)
    6.06s    = Training   runtime
    0.04s    = Validation runtime
Fitting model: CatBoost ...
    0.6804   = Validation score   (accuracy)
    6.64s    = Training   runtime
    0.01s    = Validation runtime
Fitting model: ExtraTreesGini ...
    0.6558   = Validation score   (accuracy)
    0.8s     = Training   runtime
    0.04s    = Validation runtime
Fitting model: ExtraTreesEntr ...
    0.6533   = Validation score   (accuracy)
    0.86s    = Training   runtime
    0.04s    = Validation runtime
Fitting model: NeuralNetFastAI ...
No improvement since epoch 7: early stopping
    0.7012   = Validation score   (accuracy)
    11.36s   = Training   runtime
    0.02s    = Validation runtime
Fitting model: XGBoost ...
    0.6866   = Validation score   (accuracy)
    2.02s    = Training   runtime
    0.01s    = Validation runtime
Fitting model: NeuralNetTorch ...
    0.707    = Validation score   (accuracy)
    7.38s    = Training   runtime
    0.09s    = Validation runtime
Fitting model: LightGBMLarge ...
    0.6758   = Validation score   (accuracy)
    3.62s    = Training   runtime
    0.01s    = Validation runtime
Fitting model: WeightedEnsemble_L2 ...
    0.7179   = Validation score   (accuracy)
    0.82s    = Training   runtime
    0.0s     = Validation runtime
AutoGluon training complete, total runtime = 49.84s ... Best model: "WeightedEnsemble_L2"
TabularPredictor saved. To load, use: predictor = TabularPredictor.load("AutogluonModels/ag-20240125_002023/")
Exception ignored on calling ctypes callback function: <function _ThreadpoolInfo._find_modules_with_dl_iterate_phdr.<locals>.match_module_callback at 0x7fcd700e30d0>
Traceback (most recent call last):
  File "/home/coco/anaconda3/envs/py38/lib/python3.8/site-packages/threadpoolctl.py", line 400, in match_module_callback
    self._make_module_from_path(filepath)
  File "/home/coco/anaconda3/envs/py38/lib/python3.8/site-packages/threadpoolctl.py", line 515, in _make_module_from_path
    module = module_class(filepath, prefix, user_api, internal_api)
  File "/home/coco/anaconda3/envs/py38/lib/python3.8/site-packages/threadpoolctl.py", line 606, in __init__
    self.version = self.get_version()
  File "/home/coco/anaconda3/envs/py38/lib/python3.8/site-packages/threadpoolctl.py", line 646, in get_version
    config = get_config().split()
AttributeError: 'NoneType' object has no attribute 'split'

	accuracy_score	precision_score	recall_score	f1_score	roc_auc_score
0	0.599234	0.0	0.0	0.0	0.499584

---

0.4 / 0.3

df = pd.read_csv("~/Desktop/fraudTrain.csv")

df = throw(df, 0.4)
df_tr, df_tst = split_dataframe(df, 0.3)
df_, mask = concat(df_tr, df_tst)
G_down = build_graph_tripartite(df_)
train_edges, test_edges, train_labels, test_labels = train_test_split(list(range(len(G_down.edges))), 
                                                                      list(nx.get_edge_attributes(G_down, "label").values()), 
                                                                      test_size=0.20, 
                                                                      random_state=42)
edgs = list(G_down.edges)
train_graph = G_down.edge_subgraph([edgs[x] for x in train_edges]).copy()      
train_graph.add_nodes_from(list(set(G_down.nodes) - set(train_graph.nodes)))


node2vec_train = Node2Vec(train_graph, weight_key='weight')
model_train = node2vec_train.fit(window=10)
classes = [HadamardEmbedder, AverageEmbedder, WeightedL1Embedder, WeightedL2Embedder]
for cl in classes:
    embeddings_train = cl(keyed_vectors=model_train.wv) 

train_embeddings = [embeddings_train[str(edgs[x][0]), str(edgs[x][1])] for x in train_edges]
test_embeddings = [embeddings_train[str(edgs[x][0]), str(edgs[x][1])] for x in test_edges]


# DataFrame 생성
columns = [f'X_{i}' for i in range(np.array(train_embeddings).shape[1])]
df_data = pd.DataFrame(data=train_embeddings, columns=columns)
df_labels = pd.DataFrame(data=train_labels, columns=['label'])

df = pd.concat([df_data, df_labels], axis=1)
label = np.array(train_labels)

predictr = TabularPredictor(label='label')

predictr.fit(df) 

test = np.array(test_embeddings)

columns = [f'X_{i}' for i in range(test.shape[1])]

# DataFrame 생성
test_df = pd.DataFrame(data=test, columns=columns)

y = np.array(test_labels)
yhat = predictr.predict(test_df)

evaluation(y,yhat)

Generating walks (CPU: 1): 100%|██████████| 10/10 [00:32<00:00,  3.28s/it]
No path specified. Models will be saved in: "AutogluonModels/ag-20240125_002828/"
Beginning AutoGluon training ...
AutoGluon will save models to "AutogluonModels/ag-20240125_002828/"
AutoGluon Version:  0.8.2
Python Version:     3.8.18
Operating System:   Linux
Platform Machine:   x86_64
Platform Version:   #38~22.04.1-Ubuntu SMP PREEMPT_DYNAMIC Thu Nov  2 18:01:13 UTC 2
Disk Space Avail:   619.33 GB / 982.82 GB (63.0%)
Train Data Rows:    24024
Train Data Columns: 128
Label Column: label
Preprocessing data ...
AutoGluon infers your prediction problem is: 'binary' (because only two unique label-values observed).
    2 unique label values:  [1, 0]
    If 'binary' is not the correct problem_type, please manually specify the problem_type parameter during predictor init (You may specify problem_type as one of: ['binary', 'multiclass', 'regression'])
Selected class <--> label mapping:  class 1 = 1, class 0 = 0
Using Feature Generators to preprocess the data ...
Fitting AutoMLPipelineFeatureGenerator...
    Available Memory:                    24266.65 MB
    Train Data (Original)  Memory Usage: 12.3 MB (0.1% of available memory)
    Inferring data type of each feature based on column values. Set feature_metadata_in to manually specify special dtypes of the features.
    Stage 1 Generators:
        Fitting AsTypeFeatureGenerator...
    Stage 2 Generators:
        Fitting FillNaFeatureGenerator...
    Stage 3 Generators:
        Fitting IdentityFeatureGenerator...
    Stage 4 Generators:
        Fitting DropUniqueFeatureGenerator...
    Stage 5 Generators:
        Fitting DropDuplicatesFeatureGenerator...
    Types of features in original data (raw dtype, special dtypes):
        ('float', []) : 128 | ['X_0', 'X_1', 'X_2', 'X_3', 'X_4', ...]
    Types of features in processed data (raw dtype, special dtypes):
        ('float', []) : 128 | ['X_0', 'X_1', 'X_2', 'X_3', 'X_4', ...]
    0.4s = Fit runtime
    128 features in original data used to generate 128 features in processed data.
    Train Data (Processed) Memory Usage: 12.3 MB (0.1% of available memory)
Data preprocessing and feature engineering runtime = 0.42s ...
AutoGluon will gauge predictive performance using evaluation metric: 'accuracy'
    To change this, specify the eval_metric parameter of Predictor()
Automatically generating train/validation split with holdout_frac=0.1, Train Rows: 21621, Val Rows: 2403
User-specified model hyperparameters to be fit:
{
    'NN_TORCH': {},
    'GBM': [{'extra_trees': True, 'ag_args': {'name_suffix': 'XT'}}, {}, 'GBMLarge'],
    'CAT': {},
    'XGB': {},
    'FASTAI': {},
    'RF': [{'criterion': 'gini', 'ag_args': {'name_suffix': 'Gini', 'problem_types': ['binary', 'multiclass']}}, {'criterion': 'entropy', 'ag_args': {'name_suffix': 'Entr', 'problem_types': ['binary', 'multiclass']}}, {'criterion': 'squared_error', 'ag_args': {'name_suffix': 'MSE', 'problem_types': ['regression', 'quantile']}}],
    'XT': [{'criterion': 'gini', 'ag_args': {'name_suffix': 'Gini', 'problem_types': ['binary', 'multiclass']}}, {'criterion': 'entropy', 'ag_args': {'name_suffix': 'Entr', 'problem_types': ['binary', 'multiclass']}}, {'criterion': 'squared_error', 'ag_args': {'name_suffix': 'MSE', 'problem_types': ['regression', 'quantile']}}],
    'KNN': [{'weights': 'uniform', 'ag_args': {'name_suffix': 'Unif'}}, {'weights': 'distance', 'ag_args': {'name_suffix': 'Dist'}}],
}
Fitting 13 L1 models ...
Fitting model: KNeighborsUnif ...
Exception ignored on calling ctypes callback function: <function _ThreadpoolInfo._find_modules_with_dl_iterate_phdr.<locals>.match_module_callback at 0x7fcb208ff4c0>
Traceback (most recent call last):
  File "/home/coco/anaconda3/envs/py38/lib/python3.8/site-packages/threadpoolctl.py", line 400, in match_module_callback
    self._make_module_from_path(filepath)
  File "/home/coco/anaconda3/envs/py38/lib/python3.8/site-packages/threadpoolctl.py", line 515, in _make_module_from_path
    module = module_class(filepath, prefix, user_api, internal_api)
  File "/home/coco/anaconda3/envs/py38/lib/python3.8/site-packages/threadpoolctl.py", line 606, in __init__
    self.version = self.get_version()
  File "/home/coco/anaconda3/envs/py38/lib/python3.8/site-packages/threadpoolctl.py", line 646, in get_version
    config = get_config().split()
AttributeError: 'NoneType' object has no attribute 'split'
    0.5389   = Validation score   (accuracy)
    0.06s    = Training   runtime
    0.39s    = Validation runtime
Fitting model: KNeighborsDist ...
Exception ignored on calling ctypes callback function: <function _ThreadpoolInfo._find_modules_with_dl_iterate_phdr.<locals>.match_module_callback at 0x7fcb208ff4c0>
Traceback (most recent call last):
  File "/home/coco/anaconda3/envs/py38/lib/python3.8/site-packages/threadpoolctl.py", line 400, in match_module_callback
    self._make_module_from_path(filepath)
  File "/home/coco/anaconda3/envs/py38/lib/python3.8/site-packages/threadpoolctl.py", line 515, in _make_module_from_path
    module = module_class(filepath, prefix, user_api, internal_api)
  File "/home/coco/anaconda3/envs/py38/lib/python3.8/site-packages/threadpoolctl.py", line 606, in __init__
    self.version = self.get_version()
  File "/home/coco/anaconda3/envs/py38/lib/python3.8/site-packages/threadpoolctl.py", line 646, in get_version
    config = get_config().split()
AttributeError: 'NoneType' object has no attribute 'split'
    0.5397   = Validation score   (accuracy)
    0.07s    = Training   runtime
    0.45s    = Validation runtime
Fitting model: LightGBMXT ...
    0.6833   = Validation score   (accuracy)
    1.88s    = Training   runtime
    0.01s    = Validation runtime
Fitting model: LightGBM ...
    0.6758   = Validation score   (accuracy)
    1.16s    = Training   runtime
    0.01s    = Validation runtime
Fitting model: RandomForestGini ...
    0.6629   = Validation score   (accuracy)
    4.52s    = Training   runtime
    0.04s    = Validation runtime
Fitting model: RandomForestEntr ...
    0.6633   = Validation score   (accuracy)
    6.25s    = Training   runtime
    0.04s    = Validation runtime
Fitting model: CatBoost ...
    0.6733   = Validation score   (accuracy)
    2.82s    = Training   runtime
    0.01s    = Validation runtime
Fitting model: ExtraTreesGini ...
    0.6604   = Validation score   (accuracy)
    0.81s    = Training   runtime
    0.04s    = Validation runtime
Fitting model: ExtraTreesEntr ...
    0.6579   = Validation score   (accuracy)
    0.85s    = Training   runtime
    0.04s    = Validation runtime
Fitting model: NeuralNetFastAI ...
No improvement since epoch 8: early stopping
    0.6991   = Validation score   (accuracy)
    11.81s   = Training   runtime
    0.02s    = Validation runtime
Fitting model: XGBoost ...
    0.6779   = Validation score   (accuracy)
    2.21s    = Training   runtime
    0.01s    = Validation runtime
Fitting model: NeuralNetTorch ...
    0.6937   = Validation score   (accuracy)
    9.84s    = Training   runtime
    0.11s    = Validation runtime
Fitting model: LightGBMLarge ...
    0.6821   = Validation score   (accuracy)
    4.86s    = Training   runtime
    0.02s    = Validation runtime
Fitting model: WeightedEnsemble_L2 ...
    0.7116   = Validation score   (accuracy)
    0.84s    = Training   runtime
    0.0s     = Validation runtime
AutoGluon training complete, total runtime = 50.43s ... Best model: "WeightedEnsemble_L2"
TabularPredictor saved. To load, use: predictor = TabularPredictor.load("AutogluonModels/ag-20240125_002828/")
Exception ignored on calling ctypes callback function: <function _ThreadpoolInfo._find_modules_with_dl_iterate_phdr.<locals>.match_module_callback at 0x7fcd700e3af0>
Traceback (most recent call last):
  File "/home/coco/anaconda3/envs/py38/lib/python3.8/site-packages/threadpoolctl.py", line 400, in match_module_callback
    self._make_module_from_path(filepath)
  File "/home/coco/anaconda3/envs/py38/lib/python3.8/site-packages/threadpoolctl.py", line 515, in _make_module_from_path
    module = module_class(filepath, prefix, user_api, internal_api)
  File "/home/coco/anaconda3/envs/py38/lib/python3.8/site-packages/threadpoolctl.py", line 606, in __init__
    self.version = self.get_version()
  File "/home/coco/anaconda3/envs/py38/lib/python3.8/site-packages/threadpoolctl.py", line 646, in get_version
    config = get_config().split()
AttributeError: 'NoneType' object has no attribute 'split'

	accuracy_score	precision_score	recall_score	f1_score	roc_auc_score
0	0.594406	0.464115	0.040066	0.073764	0.504412

---

0.5 / 0.3

df = pd.read_csv("~/Desktop/fraudTrain.csv")

df = throw(df, 0.5)
df_tr, df_tst = split_dataframe(df, 0.3)
df_, mask = concat(df_tr, df_tst)
G_down = build_graph_tripartite(df_)
train_edges, test_edges, train_labels, test_labels = train_test_split(list(range(len(G_down.edges))), 
                                                                      list(nx.get_edge_attributes(G_down, "label").values()), 
                                                                      test_size=0.20, 
                                                                      random_state=42)
edgs = list(G_down.edges)
train_graph = G_down.edge_subgraph([edgs[x] for x in train_edges]).copy()      
train_graph.add_nodes_from(list(set(G_down.nodes) - set(train_graph.nodes)))


node2vec_train = Node2Vec(train_graph, weight_key='weight')
model_train = node2vec_train.fit(window=10)
classes = [HadamardEmbedder, AverageEmbedder, WeightedL1Embedder, WeightedL2Embedder]
for cl in classes:
    embeddings_train = cl(keyed_vectors=model_train.wv) 

train_embeddings = [embeddings_train[str(edgs[x][0]), str(edgs[x][1])] for x in train_edges]
test_embeddings = [embeddings_train[str(edgs[x][0]), str(edgs[x][1])] for x in test_edges]


# DataFrame 생성
columns = [f'X_{i}' for i in range(np.array(train_embeddings).shape[1])]
df_data = pd.DataFrame(data=train_embeddings, columns=columns)
df_labels = pd.DataFrame(data=train_labels, columns=['label'])

df = pd.concat([df_data, df_labels], axis=1)
label = np.array(train_labels)

predictr = TabularPredictor(label='label')

predictr.fit(df) 

test = np.array(test_embeddings)

columns = [f'X_{i}' for i in range(test.shape[1])]

# DataFrame 생성
test_df = pd.DataFrame(data=test, columns=columns)

y = np.array(test_labels)
yhat = predictr.predict(test_df)

evaluation(y,yhat)

Generating walks (CPU: 1): 100%|██████████| 10/10 [00:25<00:00,  2.54s/it]
No path specified. Models will be saved in: "AutogluonModels/ag-20240125_004947/"
Beginning AutoGluon training ...
AutoGluon will save models to "AutogluonModels/ag-20240125_004947/"
AutoGluon Version:  0.8.2
Python Version:     3.8.18
Operating System:   Linux
Platform Machine:   x86_64
Platform Version:   #38~22.04.1-Ubuntu SMP PREEMPT_DYNAMIC Thu Nov  2 18:01:13 UTC 2
Disk Space Avail:   617.43 GB / 982.82 GB (62.8%)
Train Data Rows:    19219
Train Data Columns: 128
Label Column: label
Preprocessing data ...
AutoGluon infers your prediction problem is: 'binary' (because only two unique label-values observed).
    2 unique label values:  [0, 1]
    If 'binary' is not the correct problem_type, please manually specify the problem_type parameter during predictor init (You may specify problem_type as one of: ['binary', 'multiclass', 'regression'])
Selected class <--> label mapping:  class 1 = 1, class 0 = 0
Using Feature Generators to preprocess the data ...
Fitting AutoMLPipelineFeatureGenerator...
    Available Memory:                    24477.85 MB
    Train Data (Original)  Memory Usage: 9.84 MB (0.0% of available memory)
    Inferring data type of each feature based on column values. Set feature_metadata_in to manually specify special dtypes of the features.
    Stage 1 Generators:
        Fitting AsTypeFeatureGenerator...
    Stage 2 Generators:
        Fitting FillNaFeatureGenerator...
    Stage 3 Generators:
        Fitting IdentityFeatureGenerator...
    Stage 4 Generators:
        Fitting DropUniqueFeatureGenerator...
    Stage 5 Generators:
        Fitting DropDuplicatesFeatureGenerator...
    Types of features in original data (raw dtype, special dtypes):
        ('float', []) : 128 | ['X_0', 'X_1', 'X_2', 'X_3', 'X_4', ...]
    Types of features in processed data (raw dtype, special dtypes):
        ('float', []) : 128 | ['X_0', 'X_1', 'X_2', 'X_3', 'X_4', ...]
    0.7s = Fit runtime
    128 features in original data used to generate 128 features in processed data.
    Train Data (Processed) Memory Usage: 9.84 MB (0.0% of available memory)
Data preprocessing and feature engineering runtime = 0.69s ...
AutoGluon will gauge predictive performance using evaluation metric: 'accuracy'
    To change this, specify the eval_metric parameter of Predictor()
Automatically generating train/validation split with holdout_frac=0.1, Train Rows: 17297, Val Rows: 1922
User-specified model hyperparameters to be fit:
{
    'NN_TORCH': {},
    'GBM': [{'extra_trees': True, 'ag_args': {'name_suffix': 'XT'}}, {}, 'GBMLarge'],
    'CAT': {},
    'XGB': {},
    'FASTAI': {},
    'RF': [{'criterion': 'gini', 'ag_args': {'name_suffix': 'Gini', 'problem_types': ['binary', 'multiclass']}}, {'criterion': 'entropy', 'ag_args': {'name_suffix': 'Entr', 'problem_types': ['binary', 'multiclass']}}, {'criterion': 'squared_error', 'ag_args': {'name_suffix': 'MSE', 'problem_types': ['regression', 'quantile']}}],
    'XT': [{'criterion': 'gini', 'ag_args': {'name_suffix': 'Gini', 'problem_types': ['binary', 'multiclass']}}, {'criterion': 'entropy', 'ag_args': {'name_suffix': 'Entr', 'problem_types': ['binary', 'multiclass']}}, {'criterion': 'squared_error', 'ag_args': {'name_suffix': 'MSE', 'problem_types': ['regression', 'quantile']}}],
    'KNN': [{'weights': 'uniform', 'ag_args': {'name_suffix': 'Unif'}}, {'weights': 'distance', 'ag_args': {'name_suffix': 'Dist'}}],
}
Fitting 13 L1 models ...
Fitting model: KNeighborsUnif ...
Exception ignored on calling ctypes callback function: <function _ThreadpoolInfo._find_modules_with_dl_iterate_phdr.<locals>.match_module_callback at 0x7fcd8f2fa700>
Traceback (most recent call last):
  File "/home/coco/anaconda3/envs/py38/lib/python3.8/site-packages/threadpoolctl.py", line 400, in match_module_callback
    self._make_module_from_path(filepath)
  File "/home/coco/anaconda3/envs/py38/lib/python3.8/site-packages/threadpoolctl.py", line 515, in _make_module_from_path
    module = module_class(filepath, prefix, user_api, internal_api)
  File "/home/coco/anaconda3/envs/py38/lib/python3.8/site-packages/threadpoolctl.py", line 606, in __init__
    self.version = self.get_version()
  File "/home/coco/anaconda3/envs/py38/lib/python3.8/site-packages/threadpoolctl.py", line 646, in get_version
    config = get_config().split()
AttributeError: 'NoneType' object has no attribute 'split'
    0.5484   = Validation score   (accuracy)
    0.05s    = Training   runtime
    0.3s     = Validation runtime
Fitting model: KNeighborsDist ...
Exception ignored on calling ctypes callback function: <function _ThreadpoolInfo._find_modules_with_dl_iterate_phdr.<locals>.match_module_callback at 0x7fcd8f2fa700>
Traceback (most recent call last):
  File "/home/coco/anaconda3/envs/py38/lib/python3.8/site-packages/threadpoolctl.py", line 400, in match_module_callback
    self._make_module_from_path(filepath)
  File "/home/coco/anaconda3/envs/py38/lib/python3.8/site-packages/threadpoolctl.py", line 515, in _make_module_from_path
    module = module_class(filepath, prefix, user_api, internal_api)
  File "/home/coco/anaconda3/envs/py38/lib/python3.8/site-packages/threadpoolctl.py", line 606, in __init__
    self.version = self.get_version()
  File "/home/coco/anaconda3/envs/py38/lib/python3.8/site-packages/threadpoolctl.py", line 646, in get_version
    config = get_config().split()
AttributeError: 'NoneType' object has no attribute 'split'
    0.5484   = Validation score   (accuracy)
    0.07s    = Training   runtime
    0.36s    = Validation runtime
Fitting model: LightGBMXT ...
    0.6556   = Validation score   (accuracy)
    1.15s    = Training   runtime
    0.01s    = Validation runtime
Fitting model: LightGBM ...
    0.6493   = Validation score   (accuracy)
    1.14s    = Training   runtime
    0.01s    = Validation runtime
Fitting model: RandomForestGini ...
    0.6379   = Validation score   (accuracy)
    3.71s    = Training   runtime
    0.04s    = Validation runtime
Fitting model: RandomForestEntr ...
    0.6415   = Validation score   (accuracy)
    5.18s    = Training   runtime
    0.04s    = Validation runtime
Fitting model: CatBoost ...
    0.6545   = Validation score   (accuracy)
    2.32s    = Training   runtime
    0.0s     = Validation runtime
Fitting model: ExtraTreesGini ...
    0.6332   = Validation score   (accuracy)
    0.71s    = Training   runtime
    0.04s    = Validation runtime
Fitting model: ExtraTreesEntr ...
    0.6498   = Validation score   (accuracy)
    0.73s    = Training   runtime
    0.04s    = Validation runtime
Fitting model: NeuralNetFastAI ...
No improvement since epoch 4: early stopping
    0.6597   = Validation score   (accuracy)
    9.97s    = Training   runtime
    0.02s    = Validation runtime
Fitting model: XGBoost ...
    0.629    = Validation score   (accuracy)
    1.53s    = Training   runtime
    0.01s    = Validation runtime
Fitting model: NeuralNetTorch ...
    0.6514   = Validation score   (accuracy)
    5.69s    = Training   runtime
    0.08s    = Validation runtime
Fitting model: LightGBMLarge ...
    0.6472   = Validation score   (accuracy)
    3.91s    = Training   runtime
    0.01s    = Validation runtime
Fitting model: WeightedEnsemble_L2 ...
    0.6769   = Validation score   (accuracy)
    0.75s    = Training   runtime
    0.0s     = Validation runtime
AutoGluon training complete, total runtime = 39.24s ... Best model: "WeightedEnsemble_L2"
TabularPredictor saved. To load, use: predictor = TabularPredictor.load("AutogluonModels/ag-20240125_004947/")
Exception ignored on calling ctypes callback function: <function _ThreadpoolInfo._find_modules_with_dl_iterate_phdr.<locals>.match_module_callback at 0x7fccf0308310>
Traceback (most recent call last):
  File "/home/coco/anaconda3/envs/py38/lib/python3.8/site-packages/threadpoolctl.py", line 400, in match_module_callback
    self._make_module_from_path(filepath)
  File "/home/coco/anaconda3/envs/py38/lib/python3.8/site-packages/threadpoolctl.py", line 515, in _make_module_from_path
    module = module_class(filepath, prefix, user_api, internal_api)
  File "/home/coco/anaconda3/envs/py38/lib/python3.8/site-packages/threadpoolctl.py", line 606, in __init__
    self.version = self.get_version()
  File "/home/coco/anaconda3/envs/py38/lib/python3.8/site-packages/threadpoolctl.py", line 646, in get_version
    config = get_config().split()
AttributeError: 'NoneType' object has no attribute 'split'

	accuracy_score	precision_score	recall_score	f1_score	roc_auc_score
0	0.508637	0.470588	0.00339	0.006731	0.499854

---

0.5 / 0.4

df = pd.read_csv("~/Desktop/fraudTrain.csv")

df = throw(df, 0.5)
df_tr, df_tst = split_dataframe(df, 0.4)
df_, mask = concat(df_tr, df_tst)
G_down = build_graph_tripartite(df_)
train_edges, test_edges, train_labels, test_labels = train_test_split(list(range(len(G_down.edges))), 
                                                                      list(nx.get_edge_attributes(G_down, "label").values()), 
                                                                      test_size=0.20, 
                                                                      random_state=42)
edgs = list(G_down.edges)
train_graph = G_down.edge_subgraph([edgs[x] for x in train_edges]).copy()      
train_graph.add_nodes_from(list(set(G_down.nodes) - set(train_graph.nodes)))


node2vec_train = Node2Vec(train_graph, weight_key='weight')
model_train = node2vec_train.fit(window=10)
classes = [HadamardEmbedder, AverageEmbedder, WeightedL1Embedder, WeightedL2Embedder]
for cl in classes:
    embeddings_train = cl(keyed_vectors=model_train.wv) 

train_embeddings = [embeddings_train[str(edgs[x][0]), str(edgs[x][1])] for x in train_edges]
test_embeddings = [embeddings_train[str(edgs[x][0]), str(edgs[x][1])] for x in test_edges]


# DataFrame 생성
columns = [f'X_{i}' for i in range(np.array(train_embeddings).shape[1])]
df_data = pd.DataFrame(data=train_embeddings, columns=columns)
df_labels = pd.DataFrame(data=train_labels, columns=['label'])

df = pd.concat([df_data, df_labels], axis=1)
label = np.array(train_labels)

predictr = TabularPredictor(label='label')

predictr.fit(df) 

test = np.array(test_embeddings)

columns = [f'X_{i}' for i in range(test.shape[1])]

# DataFrame 생성
test_df = pd.DataFrame(data=test, columns=columns)

y = np.array(test_labels)
yhat = predictr.predict(test_df)

evaluation(y,yhat)

Generating walks (CPU: 1): 100%|██████████| 10/10 [00:25<00:00,  2.57s/it]
No path specified. Models will be saved in: "AutogluonModels/ag-20240125_004313/"
Beginning AutoGluon training ...
AutoGluon will save models to "AutogluonModels/ag-20240125_004313/"
AutoGluon Version:  0.8.2
Python Version:     3.8.18
Operating System:   Linux
Platform Machine:   x86_64
Platform Version:   #38~22.04.1-Ubuntu SMP PREEMPT_DYNAMIC Thu Nov  2 18:01:13 UTC 2
Disk Space Avail:   617.99 GB / 982.82 GB (62.9%)
Train Data Rows:    19219
Train Data Columns: 128
Label Column: label
Preprocessing data ...
AutoGluon infers your prediction problem is: 'binary' (because only two unique label-values observed).
    2 unique label values:  [0, 1]
    If 'binary' is not the correct problem_type, please manually specify the problem_type parameter during predictor init (You may specify problem_type as one of: ['binary', 'multiclass', 'regression'])
Selected class <--> label mapping:  class 1 = 1, class 0 = 0
Using Feature Generators to preprocess the data ...
Fitting AutoMLPipelineFeatureGenerator...
    Available Memory:                    24485.97 MB
    Train Data (Original)  Memory Usage: 9.84 MB (0.0% of available memory)
    Inferring data type of each feature based on column values. Set feature_metadata_in to manually specify special dtypes of the features.
    Stage 1 Generators:
        Fitting AsTypeFeatureGenerator...
    Stage 2 Generators:
        Fitting FillNaFeatureGenerator...
    Stage 3 Generators:
        Fitting IdentityFeatureGenerator...
    Stage 4 Generators:
        Fitting DropUniqueFeatureGenerator...
    Stage 5 Generators:
        Fitting DropDuplicatesFeatureGenerator...
    Types of features in original data (raw dtype, special dtypes):
        ('float', []) : 128 | ['X_0', 'X_1', 'X_2', 'X_3', 'X_4', ...]
    Types of features in processed data (raw dtype, special dtypes):
        ('float', []) : 128 | ['X_0', 'X_1', 'X_2', 'X_3', 'X_4', ...]
    0.7s = Fit runtime
    128 features in original data used to generate 128 features in processed data.
    Train Data (Processed) Memory Usage: 9.84 MB (0.0% of available memory)
Data preprocessing and feature engineering runtime = 0.68s ...
AutoGluon will gauge predictive performance using evaluation metric: 'accuracy'
    To change this, specify the eval_metric parameter of Predictor()
Automatically generating train/validation split with holdout_frac=0.1, Train Rows: 17297, Val Rows: 1922
User-specified model hyperparameters to be fit:
{
    'NN_TORCH': {},
    'GBM': [{'extra_trees': True, 'ag_args': {'name_suffix': 'XT'}}, {}, 'GBMLarge'],
    'CAT': {},
    'XGB': {},
    'FASTAI': {},
    'RF': [{'criterion': 'gini', 'ag_args': {'name_suffix': 'Gini', 'problem_types': ['binary', 'multiclass']}}, {'criterion': 'entropy', 'ag_args': {'name_suffix': 'Entr', 'problem_types': ['binary', 'multiclass']}}, {'criterion': 'squared_error', 'ag_args': {'name_suffix': 'MSE', 'problem_types': ['regression', 'quantile']}}],
    'XT': [{'criterion': 'gini', 'ag_args': {'name_suffix': 'Gini', 'problem_types': ['binary', 'multiclass']}}, {'criterion': 'entropy', 'ag_args': {'name_suffix': 'Entr', 'problem_types': ['binary', 'multiclass']}}, {'criterion': 'squared_error', 'ag_args': {'name_suffix': 'MSE', 'problem_types': ['regression', 'quantile']}}],
    'KNN': [{'weights': 'uniform', 'ag_args': {'name_suffix': 'Unif'}}, {'weights': 'distance', 'ag_args': {'name_suffix': 'Dist'}}],
}
Fitting 13 L1 models ...
Fitting model: KNeighborsUnif ...
Exception ignored on calling ctypes callback function: <function _ThreadpoolInfo._find_modules_with_dl_iterate_phdr.<locals>.match_module_callback at 0x7fccf87d4790>
Traceback (most recent call last):
  File "/home/coco/anaconda3/envs/py38/lib/python3.8/site-packages/threadpoolctl.py", line 400, in match_module_callback
    self._make_module_from_path(filepath)
  File "/home/coco/anaconda3/envs/py38/lib/python3.8/site-packages/threadpoolctl.py", line 515, in _make_module_from_path
    module = module_class(filepath, prefix, user_api, internal_api)
  File "/home/coco/anaconda3/envs/py38/lib/python3.8/site-packages/threadpoolctl.py", line 606, in __init__
    self.version = self.get_version()
  File "/home/coco/anaconda3/envs/py38/lib/python3.8/site-packages/threadpoolctl.py", line 646, in get_version
    config = get_config().split()
AttributeError: 'NoneType' object has no attribute 'split'
    0.5375   = Validation score   (accuracy)
    0.05s    = Training   runtime
    0.3s     = Validation runtime
Fitting model: KNeighborsDist ...
Exception ignored on calling ctypes callback function: <function _ThreadpoolInfo._find_modules_with_dl_iterate_phdr.<locals>.match_module_callback at 0x7fccf87d4e50>
Traceback (most recent call last):
  File "/home/coco/anaconda3/envs/py38/lib/python3.8/site-packages/threadpoolctl.py", line 400, in match_module_callback
    self._make_module_from_path(filepath)
  File "/home/coco/anaconda3/envs/py38/lib/python3.8/site-packages/threadpoolctl.py", line 515, in _make_module_from_path
    module = module_class(filepath, prefix, user_api, internal_api)
  File "/home/coco/anaconda3/envs/py38/lib/python3.8/site-packages/threadpoolctl.py", line 606, in __init__
    self.version = self.get_version()
  File "/home/coco/anaconda3/envs/py38/lib/python3.8/site-packages/threadpoolctl.py", line 646, in get_version
    config = get_config().split()
AttributeError: 'NoneType' object has no attribute 'split'
    0.5375   = Validation score   (accuracy)
    0.05s    = Training   runtime
    0.38s    = Validation runtime
Fitting model: LightGBMXT ...
    0.6774   = Validation score   (accuracy)
    0.98s    = Training   runtime
    0.01s    = Validation runtime
Fitting model: LightGBM ...
    0.667    = Validation score   (accuracy)
    1.03s    = Training   runtime
    0.0s     = Validation runtime
Fitting model: RandomForestGini ...
    0.6592   = Validation score   (accuracy)
    3.68s    = Training   runtime
    0.04s    = Validation runtime
Fitting model: RandomForestEntr ...
    0.6623   = Validation score   (accuracy)
    5.2s     = Training   runtime
    0.04s    = Validation runtime
Fitting model: CatBoost ...
    0.6727   = Validation score   (accuracy)
    3.06s    = Training   runtime
    0.01s    = Validation runtime
Fitting model: ExtraTreesGini ...
    0.6649   = Validation score   (accuracy)
    0.7s     = Training   runtime
    0.04s    = Validation runtime
Fitting model: ExtraTreesEntr ...
    0.6686   = Validation score   (accuracy)
    0.74s    = Training   runtime
    0.04s    = Validation runtime
Fitting model: NeuralNetFastAI ...
No improvement since epoch 5: early stopping
    0.6649   = Validation score   (accuracy)
    10.27s   = Training   runtime
    0.02s    = Validation runtime
Fitting model: XGBoost ...
    0.6582   = Validation score   (accuracy)
    1.74s    = Training   runtime
    0.01s    = Validation runtime
Fitting model: NeuralNetTorch ...
    0.6727   = Validation score   (accuracy)
    5.93s    = Training   runtime
    0.08s    = Validation runtime
Fitting model: LightGBMLarge ...
    0.6681   = Validation score   (accuracy)
    3.79s    = Training   runtime
    0.01s    = Validation runtime
Fitting model: WeightedEnsemble_L2 ...
    0.6915   = Validation score   (accuracy)
    0.74s    = Training   runtime
    0.0s     = Validation runtime
AutoGluon training complete, total runtime = 40.32s ... Best model: "WeightedEnsemble_L2"
TabularPredictor saved. To load, use: predictor = TabularPredictor.load("AutogluonModels/ag-20240125_004313/")
Exception ignored on calling ctypes callback function: <function _ThreadpoolInfo._find_modules_with_dl_iterate_phdr.<locals>.match_module_callback at 0x7fcd8f2579d0>
Traceback (most recent call last):
  File "/home/coco/anaconda3/envs/py38/lib/python3.8/site-packages/threadpoolctl.py", line 400, in match_module_callback
    self._make_module_from_path(filepath)
  File "/home/coco/anaconda3/envs/py38/lib/python3.8/site-packages/threadpoolctl.py", line 515, in _make_module_from_path
    module = module_class(filepath, prefix, user_api, internal_api)
  File "/home/coco/anaconda3/envs/py38/lib/python3.8/site-packages/threadpoolctl.py", line 606, in __init__
    self.version = self.get_version()
  File "/home/coco/anaconda3/envs/py38/lib/python3.8/site-packages/threadpoolctl.py", line 646, in get_version
    config = get_config().split()
AttributeError: 'NoneType' object has no attribute 'split'
/home/coco/anaconda3/envs/py38/lib/python3.8/site-packages/sklearn/metrics/_classification.py:1344: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.
  _warn_prf(average, modifier, msg_start, len(result))

	accuracy_score	precision_score	recall_score	f1_score	roc_auc_score
0	0.501977	0.0	0.0	0.0	0.5

---

0.5 / 0.2

df = pd.read_csv("~/Desktop/fraudTrain.csv")

df = throw(df, 0.5)
df_tr, df_tst = split_dataframe(df, 0.2)
df_, mask = concat(df_tr, df_tst)
G_down = build_graph_tripartite(df_)
train_edges, test_edges, train_labels, test_labels = train_test_split(list(range(len(G_down.edges))), 
                                                                      list(nx.get_edge_attributes(G_down, "label").values()), 
                                                                      test_size=0.20, 
                                                                      random_state=42)
edgs = list(G_down.edges)
train_graph = G_down.edge_subgraph([edgs[x] for x in train_edges]).copy()      
train_graph.add_nodes_from(list(set(G_down.nodes) - set(train_graph.nodes)))


node2vec_train = Node2Vec(train_graph, weight_key='weight')
model_train = node2vec_train.fit(window=10)
classes = [HadamardEmbedder, AverageEmbedder, WeightedL1Embedder, WeightedL2Embedder]
for cl in classes:
    embeddings_train = cl(keyed_vectors=model_train.wv) 

train_embeddings = [embeddings_train[str(edgs[x][0]), str(edgs[x][1])] for x in train_edges]
test_embeddings = [embeddings_train[str(edgs[x][0]), str(edgs[x][1])] for x in test_edges]


# DataFrame 생성
columns = [f'X_{i}' for i in range(np.array(train_embeddings).shape[1])]
df_data = pd.DataFrame(data=train_embeddings, columns=columns)
df_labels = pd.DataFrame(data=train_labels, columns=['label'])

df = pd.concat([df_data, df_labels], axis=1)
label = np.array(train_labels)

predictr = TabularPredictor(label='label')

predictr.fit(df) 

test = np.array(test_embeddings)

columns = [f'X_{i}' for i in range(test.shape[1])]

# DataFrame 생성
test_df = pd.DataFrame(data=test, columns=columns)

y = np.array(test_labels)
yhat = predictr.predict(test_df)

evaluation(y,yhat)

Generating walks (CPU: 1): 100%|██████████| 10/10 [00:25<00:00,  2.54s/it]
No path specified. Models will be saved in: "AutogluonModels/ag-20240125_005617/"
Beginning AutoGluon training ...
AutoGluon will save models to "AutogluonModels/ag-20240125_005617/"
AutoGluon Version:  0.8.2
Python Version:     3.8.18
Operating System:   Linux
Platform Machine:   x86_64
Platform Version:   #38~22.04.1-Ubuntu SMP PREEMPT_DYNAMIC Thu Nov  2 18:01:13 UTC 2
Disk Space Avail:   616.86 GB / 982.82 GB (62.8%)
Train Data Rows:    19219
Train Data Columns: 128
Label Column: label
Preprocessing data ...
AutoGluon infers your prediction problem is: 'binary' (because only two unique label-values observed).
    2 unique label values:  [0, 1]
    If 'binary' is not the correct problem_type, please manually specify the problem_type parameter during predictor init (You may specify problem_type as one of: ['binary', 'multiclass', 'regression'])
Selected class <--> label mapping:  class 1 = 1, class 0 = 0
Using Feature Generators to preprocess the data ...
Fitting AutoMLPipelineFeatureGenerator...
    Available Memory:                    24455.01 MB
    Train Data (Original)  Memory Usage: 9.84 MB (0.0% of available memory)
    Inferring data type of each feature based on column values. Set feature_metadata_in to manually specify special dtypes of the features.
    Stage 1 Generators:
        Fitting AsTypeFeatureGenerator...
    Stage 2 Generators:
        Fitting FillNaFeatureGenerator...
    Stage 3 Generators:
        Fitting IdentityFeatureGenerator...
    Stage 4 Generators:
        Fitting DropUniqueFeatureGenerator...
    Stage 5 Generators:
        Fitting DropDuplicatesFeatureGenerator...
    Types of features in original data (raw dtype, special dtypes):
        ('float', []) : 128 | ['X_0', 'X_1', 'X_2', 'X_3', 'X_4', ...]
    Types of features in processed data (raw dtype, special dtypes):
        ('float', []) : 128 | ['X_0', 'X_1', 'X_2', 'X_3', 'X_4', ...]
    0.6s = Fit runtime
    128 features in original data used to generate 128 features in processed data.
    Train Data (Processed) Memory Usage: 9.84 MB (0.0% of available memory)
Data preprocessing and feature engineering runtime = 0.65s ...
AutoGluon will gauge predictive performance using evaluation metric: 'accuracy'
    To change this, specify the eval_metric parameter of Predictor()
Automatically generating train/validation split with holdout_frac=0.1, Train Rows: 17297, Val Rows: 1922
User-specified model hyperparameters to be fit:
{
    'NN_TORCH': {},
    'GBM': [{'extra_trees': True, 'ag_args': {'name_suffix': 'XT'}}, {}, 'GBMLarge'],
    'CAT': {},
    'XGB': {},
    'FASTAI': {},
    'RF': [{'criterion': 'gini', 'ag_args': {'name_suffix': 'Gini', 'problem_types': ['binary', 'multiclass']}}, {'criterion': 'entropy', 'ag_args': {'name_suffix': 'Entr', 'problem_types': ['binary', 'multiclass']}}, {'criterion': 'squared_error', 'ag_args': {'name_suffix': 'MSE', 'problem_types': ['regression', 'quantile']}}],
    'XT': [{'criterion': 'gini', 'ag_args': {'name_suffix': 'Gini', 'problem_types': ['binary', 'multiclass']}}, {'criterion': 'entropy', 'ag_args': {'name_suffix': 'Entr', 'problem_types': ['binary', 'multiclass']}}, {'criterion': 'squared_error', 'ag_args': {'name_suffix': 'MSE', 'problem_types': ['regression', 'quantile']}}],
    'KNN': [{'weights': 'uniform', 'ag_args': {'name_suffix': 'Unif'}}, {'weights': 'distance', 'ag_args': {'name_suffix': 'Dist'}}],
}
Fitting 13 L1 models ...
Fitting model: KNeighborsUnif ...
Exception ignored on calling ctypes callback function: <function _ThreadpoolInfo._find_modules_with_dl_iterate_phdr.<locals>.match_module_callback at 0x7fcb208fbca0>
Traceback (most recent call last):
  File "/home/coco/anaconda3/envs/py38/lib/python3.8/site-packages/threadpoolctl.py", line 400, in match_module_callback
    self._make_module_from_path(filepath)
  File "/home/coco/anaconda3/envs/py38/lib/python3.8/site-packages/threadpoolctl.py", line 515, in _make_module_from_path
    module = module_class(filepath, prefix, user_api, internal_api)
  File "/home/coco/anaconda3/envs/py38/lib/python3.8/site-packages/threadpoolctl.py", line 606, in __init__
    self.version = self.get_version()
  File "/home/coco/anaconda3/envs/py38/lib/python3.8/site-packages/threadpoolctl.py", line 646, in get_version
    config = get_config().split()
AttributeError: 'NoneType' object has no attribute 'split'
    0.5182   = Validation score   (accuracy)
    0.05s    = Training   runtime
    0.3s     = Validation runtime
Fitting model: KNeighborsDist ...
Exception ignored on calling ctypes callback function: <function _ThreadpoolInfo._find_modules_with_dl_iterate_phdr.<locals>.match_module_callback at 0x7fcb208fbca0>
Traceback (most recent call last):
  File "/home/coco/anaconda3/envs/py38/lib/python3.8/site-packages/threadpoolctl.py", line 400, in match_module_callback
    self._make_module_from_path(filepath)
  File "/home/coco/anaconda3/envs/py38/lib/python3.8/site-packages/threadpoolctl.py", line 515, in _make_module_from_path
    module = module_class(filepath, prefix, user_api, internal_api)
  File "/home/coco/anaconda3/envs/py38/lib/python3.8/site-packages/threadpoolctl.py", line 606, in __init__
    self.version = self.get_version()
  File "/home/coco/anaconda3/envs/py38/lib/python3.8/site-packages/threadpoolctl.py", line 646, in get_version
    config = get_config().split()
AttributeError: 'NoneType' object has no attribute 'split'
    0.5193   = Validation score   (accuracy)
    0.07s    = Training   runtime
    0.38s    = Validation runtime
Fitting model: LightGBMXT ...
    0.6348   = Validation score   (accuracy)
    0.97s    = Training   runtime
    0.01s    = Validation runtime
Fitting model: LightGBM ...
    0.6348   = Validation score   (accuracy)
    1.37s    = Training   runtime
    0.01s    = Validation runtime
Fitting model: RandomForestGini ...
    0.6254   = Validation score   (accuracy)
    3.74s    = Training   runtime
    0.04s    = Validation runtime
Fitting model: RandomForestEntr ...
    0.6337   = Validation score   (accuracy)
    5.22s    = Training   runtime
    0.04s    = Validation runtime
Fitting model: CatBoost ...
    0.6462   = Validation score   (accuracy)
    3.83s    = Training   runtime
    0.0s     = Validation runtime
Fitting model: ExtraTreesGini ...
    0.6384   = Validation score   (accuracy)
    0.69s    = Training   runtime
    0.04s    = Validation runtime
Fitting model: ExtraTreesEntr ...
    0.6446   = Validation score   (accuracy)
    0.74s    = Training   runtime
    0.04s    = Validation runtime
Fitting model: NeuralNetFastAI ...
No improvement since epoch 8: early stopping
    0.6426   = Validation score   (accuracy)
    9.51s    = Training   runtime
    0.02s    = Validation runtime
Fitting model: XGBoost ...
    0.6301   = Validation score   (accuracy)
    1.54s    = Training   runtime
    0.01s    = Validation runtime
Fitting model: NeuralNetTorch ...
    0.6415   = Validation score   (accuracy)
    5.31s    = Training   runtime
    0.08s    = Validation runtime
Fitting model: LightGBMLarge ...
    0.6379   = Validation score   (accuracy)
    2.72s    = Training   runtime
    0.01s    = Validation runtime
Fitting model: WeightedEnsemble_L2 ...
    0.6634   = Validation score   (accuracy)
    0.73s    = Training   runtime
    0.0s     = Validation runtime
AutoGluon training complete, total runtime = 38.72s ... Best model: "WeightedEnsemble_L2"
TabularPredictor saved. To load, use: predictor = TabularPredictor.load("AutogluonModels/ag-20240125_005617/")

	accuracy_score	precision_score	recall_score	f1_score	roc_auc_score
0	0.510926	0.0	0.0	0.0	0.499796