validmind 1.7.0__py3-none-any.whl → 1.8.1__py3-none-any.whl

validmind/datasets/regression/datasets/lending_club_loan_rates.csv

@@ -0,0 +1,138 @@
+DATE,loan_rate_A,loan_rate_B,loan_rate_C,loan_rate_D
+2007-08-01,7.7666666666666700,9.497692307692310,10.9475,12.267
+2007-09-01,7.841428571428570,9.276666666666670,10.829166666666700,12.436666666666700
+2007-10-01,7.83,9.433333333333330,10.825925925925900,12.737368421052600
+2007-11-01,7.779090909090910,9.467777777777780,10.967037037037000,12.609444444444400
+2007-12-01,7.695833333333330,9.3875,10.805,12.47888888888890
+2008-01-01,7.961333333333330,9.693125,11.288055555555600,13.008200000000000
+2008-02-01,8.130333333333330,10.035730337078700,11.525636363636400,13.235925925925900
+2008-03-01,8.126285714285710,10.05709677419360,11.60367816091950,13.145975609756100
+2008-04-01,8.092083333333330,10.16122807017540,11.61359375,13.258260869565200
+2008-05-01,8.151428571428570,10.077916666666700,11.634814814814800,12.974
+2008-06-01,8.2055,10.068181818181800,11.649117647058800,13.290416666666700
+2008-07-01,8.220434782608700,10.158148148148100,11.78128205128210,13.3132
+2008-08-01,8.287,10.443636363636400,11.86875,13.460833333333300
+2008-09-01,8.08,10.433333333333300,11.803333333333300,13.486666666666700
+2008-10-01,8.612857142857140,10.757333333333300,12.13025641025640,13.681764705882400
+2008-11-01,8.558787878787880,10.911111111111100,12.349166666666700,13.83516129032260
+2008-12-01,9.199166666666670,11.432987012987000,12.764520547945200,14.423714285714300
+2009-01-01,8.956065573770490,11.618840579710100,13.162727272727300,14.634807692307700
+2009-02-01,8.898615384615390,12.01972972972970,13.168222222222200,14.635967741935500
+2009-03-01,8.821111111111110,12.035806451612900,13.182000000000000,14.603529411764700
+2009-04-01,8.685268817204300,12.012857142857100,13.133461538461500,14.720821917808200
+2009-05-01,8.944868421052630,11.619561403508800,13.155242718446600,14.541750000000000
+2009-06-01,8.933404255319150,11.6745,13.11355140186920,14.507916666666700
+2009-07-01,8.965098039215690,11.591623931623900,13.06843137254900,14.669272727272700
+2009-08-01,8.472935779816510,11.800933333333300,13.355959595959600,15.127391304347800
+2009-09-01,8.443076923076920,11.84740506329110,13.55984,15.162857142857100
+2009-10-01,8.436535433070870,11.86804347826090,13.482191780821900,15.3096875
+2009-11-01,8.363513513513510,11.807222222222200,13.499345238095200,15.217190082644600
+2009-12-01,8.409424460431660,11.87390134529150,13.519559748427700,15.220645161290300
+2010-01-01,8.144044117647060,11.607019230769200,13.482960526315800,15.173942307692300
+2010-02-01,7.517205882352940,10.687883817427400,13.35064705882350,15.231326530612200
+2010-03-01,7.457604790419160,10.6644,13.337486910994800,15.195703703703700
+2010-04-01,7.498095238095240,10.649088050314500,13.424000000000000,15.150000000000000
+2010-05-01,7.4599090909090900,10.723706293706300,13.427931034482800,15.29125
+2010-06-01,7.52636752136752,11.225825242718400,13.756233766233800,15.530058823529400
+2010-07-01,7.45978102189781,11.22090909090910,13.86310606060610,15.569890710382500
+2010-08-01,7.452052631578950,11.248456973293800,13.835163636363600,15.517061855670100
+2010-09-01,7.498099173553720,11.170552325581400,13.797773279352200,15.597903225806500
+2010-10-01,6.891980519480520,10.575714285714300,13.492008368200800,15.294080459770100
+2010-11-01,6.3666201117318400,9.736618497109830,12.948222222222200,14.853270440251600
+2010-12-01,6.319193954659950,9.722658959537570,12.96254752851710,14.805481927710800
+2011-01-01,6.582382133995040,10.1541475826972,13.06860655737710,15.108478260869600
+2011-02-01,6.80948717948718,10.467368421052600,13.297196261682200,15.395933333333300
+2011-03-01,6.834241573033710,10.474313253012000,13.317777777777800,15.337074468085100
+2011-04-01,6.847228260869570,10.450381526104400,13.228860294117600,15.29950226244340
+2011-05-01,7.165149253731340,11.099935760171300,13.817382352941200,16.48267716535430
+2011-06-01,7.018872651356990,11.108265682656800,13.876910112359600,16.51079831932770
+2011-07-01,7.083340563991320,11.067542955326500,13.886373937677100,16.48506172839510
+2011-08-01,7.0728650646950100,11.086491228070200,13.819787234042600,16.492789855072500
+2011-09-01,7.084527131782950,11.254157814871000,14.151633802816900,16.680548523206800
+2011-10-01,7.351312410841660,11.598674121405800,14.406036036036000,17.25779661016950
+2011-11-01,7.4149581239531,11.555904628331000,14.424929906542100,17.232988505747100
+2011-12-01,7.589288888888890,11.589472222222200,14.50002105263160,17.223344155844200
+2012-01-01,7.525056980056980,11.635782556750300,14.443313953488400,17.08759375
+2012-02-01,7.610733695652170,11.597185929648200,14.58754756871040,17.420031746031700
+2012-03-01,7.579030100334450,11.886394763343400,14.80486220472440,18.253422818791900
+2012-04-01,7.5876,11.925473321858900,14.861176470588200,18.16556786703600
+2012-05-01,7.476845102505700,12.016376554174100,14.810244252873600,18.233905472636800
+2012-06-01,7.5403636363636400,12.15405750798720,14.900022573363400,18.17427304964540
+2012-07-01,7.6588717339667500,12.227395115842200,15.313709677419400,18.569943181818200
+2012-08-01,7.574986474301170,12.34835970464140,15.457079288025900,18.649875862069000
+2012-09-01,7.596716681376880,12.335429844098000,15.458238267148000,18.6446126340882
+2012-10-01,7.683201877934270,12.275302245250400,15.453846710050600,18.670659574468100
+2012-11-01,7.61423155737705,12.337397568662800,15.451,18.65142300194930
+2012-12-01,7.822746823069400,12.1065440464666,15.75559186189890,18.592207478890200
+2013-01-01,7.752121546961330,12.137295401403000,15.77893272635310,18.554923913043500
+2013-02-01,7.698670360110800,12.109676284306800,15.784549549549500,18.576124721603600
+2013-03-01,7.720320610687020,12.052285819793200,15.80080592105260,18.641330998248700
+2013-04-01,7.946095025983670,12.05531282405810,15.81847364568080,18.610517928286900
+2013-05-01,7.573445527015060,12.067179799209000,15.781325174825200,18.6713679245283
+2013-06-01,7.520518828451880,11.944434831147100,15.796356209150300,18.66170493685420
+2013-07-01,7.605866466616650,11.357245404868400,15.23495371752760,18.689437963944900
+2013-08-01,7.606510948905110,11.37418853255590,15.196370835609000,18.714716883116900
+2013-09-01,7.695094055680960,11.49977115613830,15.379319657231900,18.779754016064300
+2013-10-01,7.8066564417177900,11.753876332622600,15.673991344073100,18.88925198690980
+2013-11-01,7.837439117199390,11.892656587473000,15.553595594020500,18.89677792612370
+2013-12-01,7.888719008264460,11.915244541484700,15.146307539188900,18.39415185783520
+2014-01-01,7.884319131161240,11.903197773972600,15.017733510402800,18.154655041698300
+2014-02-01,7.84595130748422,11.806563223714700,14.829327827483400,18.013829250720500
+2014-03-01,7.881642651296830,11.75884096438840,14.674028332003200,17.929929601072700
+2014-04-01,7.886479457218240,11.759009712435700,14.685187592867800,17.975035360678900
+2014-05-01,7.529020283199390,11.264750000000000,14.191762042738100,17.158339091150700
+2014-06-01,7.351805118994160,11.008507223114000,13.938348755912000,16.76545196134170
+2014-07-01,7.351720283533260,10.991443699731900,13.962106592877800,16.78448344701380
+2014-08-01,7.305786460087570,10.92469630557300,13.931857451403900,16.821789838337200
+2014-09-01,7.1805243268776600,10.92935610522180,13.932417061611400,16.81521568627450
+2014-10-01,7.313578708946770,10.99294086810650,13.902686392661200,16.741794034723300
+2014-11-01,7.341760975609760,10.609953738248000,13.599219115824300,16.518561545801500
+2014-12-01,7.2349226160758900,10.451943852167700,13.558856439127400,16.457166246851400
+2015-01-01,7.284862446138550,10.523150939274800,13.580668403647700,16.49141867927720
+2015-02-01,7.126818078064370,10.230664872566100,13.3875424992544,16.665386680988200
+2015-03-01,7.053031877213700,10.09241054613940,13.314671177266600,16.752682622787300
+2015-04-01,7.028140703517590,10.05959205467750,13.339224530168200,16.770982191041600
+2015-05-01,6.942790697674420,10.075693915181300,13.309596621846900,16.8081057178116
+2015-06-01,6.924908338261380,10.016840254294600,13.311009585460000,16.775532012897300
+2015-07-01,6.846263779016990,9.973752109959010,13.315842231914600,16.76153384747220
+2015-08-01,6.895340965654560,9.938369829683700,13.294216701173200,16.7434092634776
+2015-09-01,6.883837661010020,9.956684323369910,13.28493008739080,16.776011560693600
+2015-10-01,6.857467413674820,9.9480314478252,13.308447903156100,16.75174123337360
+2015-11-01,6.818325145839730,10.012865733184400,13.114165689017200,16.679007398273700
+2015-12-01,6.813621571972180,9.920643664396380,13.113938948995400,16.720045469308200
+2016-01-01,6.854723867456380,9.953370355808930,13.313586302637700,16.981329494896600
+2016-02-01,6.844791755508170,9.962073382887870,13.51001321003960,17.7306925087108
+2016-03-01,6.51158106060606,9.981660457385230,13.518506973417600,17.82086815920400
+2016-04-01,6.459687188434700,10.000419839473100,13.54169860126460,17.77998596913210
+2016-05-01,6.905898270251760,9.948223350253810,13.501587610160700,17.805547243003600
+2016-06-01,7.12222120518688,10.17323210606470,13.780876188418300,18.284878048780500
+2016-07-01,7.409693532818530,10.48121790772040,14.012340662161000,18.673088637251100
+2016-08-01,7.415675891431610,10.430686851479000,14.018968108344300,18.725051626729000
+2016-09-01,7.309269304403320,10.420682593856700,13.99552615321250,18.80164383561640
+2016-10-01,7.155384800384800,10.461156528477300,14.06901384239570,18.808550010269000
+2016-11-01,7.026321948889420,10.626164935530100,14.117570806100200,18.7734665882814
+2016-12-01,6.989425942156000,10.574707292338900,14.143388450148100,18.831436605317000
+2017-01-01,6.922970447631470,10.593491575712600,14.140238904713800,18.479066305818700
+2017-02-01,6.9446164739202300,10.640543691703500,14.111862055097500,18.06909398034400
+2017-03-01,6.893785237086530,10.600121526226400,14.140748368044800,18.110767457939600
+2017-04-01,6.89110229645094,10.622533764284900,14.136295552367300,18.084069239500600
+2017-05-01,6.936799809795530,10.557189495458200,14.272662721893500,18.665688052068800
+2017-06-01,7.015996470069130,10.566750889988700,14.37513654096230,19.046772499403200
+2017-07-01,7.0019612448620100,10.566760062456600,14.345889606150700,19.107410634495100
+2017-08-01,6.985308885383810,10.576519034772200,14.36054723695700,19.087508
+2017-09-01,7.08636862745098,10.51879065667700,14.307342195158200,19.013268023587600
+2017-10-01,7.113857827476040,10.522347111638200,14.222900293532800,18.95540403155130
+2017-11-01,6.925136533803130,10.513319831983200,14.159044066620400,18.923875141517100
+2017-12-01,6.8353083235638900,10.52431454716500,14.15896581775930,18.952280408231600
+2018-01-01,6.77865500934785,10.516662401754700,14.194534883720900,18.959176493330800
+2018-02-01,6.770555486361940,10.53045831202050,14.188890264490700,19.05002990175140
+2018-03-01,6.715374727894680,10.540809179770500,14.160395894428200,19.323733236680000
+2018-04-01,6.704338970023060,10.562063789868700,14.168836298629400,19.367843168957200
+2018-05-01,6.768701702826520,10.630962232202000,14.457164179104500,19.322808271248300
+2018-06-01,6.812492979436540,10.706152809367400,14.628350747613000,19.307152249134900
+2018-07-01,7.1668788793826500,11.116104527240200,15.085822900484300,19.769436012142500
+2018-08-01,7.218996976568410,11.161285669241700,15.142618464766600,19.85760335716510
+2018-09-01,7.2012805625971100,11.191917953668000,15.139769364664900,19.74845895522390
+2018-10-01,7.228498120038990,11.208417521704800,15.129104518736200,19.79216299416720
+2018-11-01,7.536896956485290,11.390483083511800,15.12686942339060,19.632696590118300
+2018-12-01,7.715209008701590,11.45963094824590,15.107476048212600,19.558346273291900

validmind/datasets/regression/fred.py

@@ -0,0 +1,132 @@
+import os
+import pickle
+import pandas as pd
+
+current_path = os.path.dirname(os.path.abspath(__file__))
+dataset_path = os.path.join(current_path, "datasets")
+models_path = os.path.join(current_path, "models")
+
+
+target_column = "MORTGAGE30US"
+feature_columns = ["FEDFUNDS", "GS10", "UNRATE"]
+frequency = "MS"
+split_option = "train_test"
+transform_func = "diff"
+
+
+def load_data():
+    data_file = os.path.join(dataset_path, "fred_loan_rates.csv")
+    df = pd.read_csv(data_file, parse_dates=["DATE"], index_col="DATE")
+    df = df[[target_column] + feature_columns]
+    return df
+
+
+def preprocess(df, split_option="train_test_val", train_size=0.6, test_size=0.2):
+    """
+    Split a time series DataFrame into train, validation, and test sets.
+
+    Parameters:
+        df (pandas.DataFrame): The time series DataFrame to be split.
+        split_option (str): The split option to choose from: 'train_test_val' (default) or 'train_test'.
+        train_size (float): The proportion of the dataset to include in the training set. Default is 0.6.
+        test_size (float): The proportion of the dataset to include in the test set. Default is 0.2.
+
+    Returns:
+        train_df (pandas.DataFrame): The training set.
+        validation_df (pandas.DataFrame): The validation set (only returned if split_option is 'train_test_val').
+        test_df (pandas.DataFrame): The test set.
+
+    """
+    # Sort the DataFrame by the time column (assuming the time column is the index)
+    df = df.sort_index()
+
+    if split_option == "train_test_val":
+        # Split the DataFrame into train, validation, and test sets
+        train_size = int(len(df) * train_size)
+        val_size = int(len(df) * test_size)
+
+        train_df = df.iloc[:train_size]
+        validation_df = df.iloc[train_size : train_size + val_size]
+        test_df = df.iloc[train_size + val_size :]
+
+        return train_df, validation_df, test_df
+
+    elif split_option == "train_test":
+        # Split the DataFrame into train and test sets
+        train_size = int(len(df) * train_size)
+
+        train_df = df.iloc[:train_size]
+        test_df = df.iloc[train_size:]
+
+        return train_df, test_df
+
+    else:
+        raise ValueError(
+            "Invalid split_option. Must be 'train_test_val' or 'train_test'."
+        )
+
+
+def transform(df, transform_func="diff"):
+    if transform_func == "diff":
+        df = df.diff().dropna()
+    return df
+
+
+def load_model(model_name):
+    model_file = model_name + ".pkl"
+    model_path = os.path.join(models_path, model_file)
+
+    if os.path.isfile(model_path):
+        with open(model_path, "rb") as f:
+            model = pickle.load(f)
+
+        train_df = load_train_dataset(model_path)
+        test_df = load_test_dataset(model_name)
+
+        return model, train_df, test_df
+    else:
+        print(f"No model file found with the name: {model_name}")
+        return None, None, None
+
+
+def load_train_dataset(model_path):
+    with open(model_path, "rb") as f:
+        model_fit = pickle.load(f)
+
+    # Extract the endogenous (target) variable from the model
+    train_df = pd.Series(model_fit.model.endog, index=model_fit.model.data.row_labels)
+    train_df = train_df.to_frame()
+    target_var_name = model_fit.model.endog_names
+    train_df.columns = [target_var_name]
+
+    # Extract the exogenous (explanatory) variables from the model
+    exog_df = pd.DataFrame(
+        model_fit.model.exog,
+        index=model_fit.model.data.row_labels,
+        columns=model_fit.model.exog_names,
+    )
+
+    # Concatenate the endogenous (target) and exogenous (explanatory) variables
+    train_df = pd.concat([train_df, exog_df], axis=1)
+
+    return train_df
+
+
+def load_test_dataset(model_name):
+    if model_name == "fred_loan_rates_model_1":
+        filename = "fred_loan_rates_test_1.csv"
+    elif model_name == "fred_loan_rates_model_2":
+        filename = "fred_loan_rates_test_2.csv"
+    elif model_name == "fred_loan_rates_model_3":
+        filename = "fred_loan_rates_test_3.csv"
+    elif model_name == "fred_loan_rates_model_4":
+        filename = "fred_loan_rates_test_4.csv"
+    elif model_name == "fred_loan_rates_model_5":
+        filename = "fred_loan_rates_test_5.csv"
+    else:
+        return None
+
+    data_file = os.path.join(dataset_path, filename)
+    df = pd.read_csv(data_file, parse_dates=["DATE"], index_col="DATE")
+    df = df.diff().dropna()
+    return df

validmind/datasets/regression/lending_club.py

@@ -0,0 +1,70 @@
+import os
+
+import pandas as pd
+
+current_path = os.path.dirname(os.path.abspath(__file__))
+dataset_path = os.path.join(
+    current_path, "..", "..", "..", "notebooks", "datasets", "time_series"
+)
+
+target_column = ["loan_rate_A"]
+feature_columns = ["loan_rate_B", "loan_rate_C", "loan_rate_D"]
+frequency = "MS"
+split_option = "train_test"
+
+
+def load_data():
+    data_file = os.path.join(dataset_path, "lending_club_loan_rates.csv")
+    df = pd.read_csv(data_file, parse_dates=["DATE"], index_col="DATE")
+    return df
+
+
+def preprocess(df, split_option="train_test_val", train_size=0.6, test_size=0.2):
+    """
+    Split a time series DataFrame into train, validation, and test sets.
+
+    Parameters:
+        df (pandas.DataFrame): The time series DataFrame to be split.
+        split_option (str): The split option to choose from: 'train_test_val' (default) or 'train_test'.
+        train_size (float): The proportion of the dataset to include in the training set. Default is 0.6.
+        test_size (float): The proportion of the dataset to include in the test set. Default is 0.2.
+
+    Returns:
+        train_df (pandas.DataFrame): The training set.
+        validation_df (pandas.DataFrame): The validation set (only returned if split_option is 'train_test_val').
+        test_df (pandas.DataFrame): The test set.
+
+    """
+    # Sort the DataFrame by the time column (assuming the time column is the index)
+    df = df.sort_index()
+
+    if split_option == "train_test_val":
+        # Split the DataFrame into train, validation, and test sets
+        train_size = int(len(df) * train_size)
+        val_size = int(len(df) * test_size)
+
+        train_df = df.iloc[:train_size]
+        validation_df = df.iloc[train_size : train_size + val_size]
+        test_df = df.iloc[train_size + val_size :]
+
+        return train_df, validation_df, test_df
+
+    elif split_option == "train_test":
+        # Split the DataFrame into train and test sets
+        train_size = int(len(df) * train_size)
+
+        train_df = df.iloc[:train_size]
+        test_df = df.iloc[train_size:]
+
+        return train_df, test_df
+
+    else:
+        raise ValueError(
+            "Invalid split_option. Must be 'train_test_val' or 'train_test'."
+        )
+
+
+def transform(df, transform_func="diff"):
+    if transform_func == "diff":
+        df = df.diff().dropna()
+    return df

validmind/model_validation/sklearn/{threshold_tests.pyx → threshold_tests.py}

@@ -203,12 +203,12 @@ class OverfitDiagnosis(ThresholdTest):
         prediction_column = f"{target_column}_pred"
 
         # Add prediction column in the training dataset
-        train_df = self.model.train_ds.df.copy(deep=True)
+        train_df = self.model.train_ds.df.copy()
         train_class_pred = self.model.class_predictions(self.model.y_train_predict)
         train_df[prediction_column] = train_class_pred
 
         # Add prediction column in the test dataset
-        test_df = self.model.test_ds.df.copy(deep=True)
+        test_df = self.model.test_ds.df.copy()
         test_class_pred = self.model.class_predictions(self.model.y_test_predict)
         test_df[prediction_column] = test_class_pred
 
@@ -293,7 +293,7 @@ class OverfitDiagnosis(ThresholdTest):
 
         results_train = pd.DataFrame(results_train)
         results_test = pd.DataFrame(results_test)
-        results = results_train.copy(deep=True)
+        results = results_train.copy()
         results.rename(
             columns={"shape": "training records", "accuracy": "training accuracy"},
             inplace=True,
@@ -458,11 +458,11 @@ class WeakspotsDiagnosis(ThresholdTest):
         target_column = self.model.train_ds.target_column
         prediction_column = f"{target_column}_pred"
 
-        train_df = self.model.train_ds.df.copy(deep=True)
+        train_df = self.model.train_ds.df.copy()
         train_class_pred = self.model.class_predictions(self.model.y_train_predict)
         train_df[prediction_column] = train_class_pred
 
-        test_df = self.model.test_ds.df.copy(deep=True)
+        test_df = self.model.test_ds.df.copy()
         test_class_pred = self.model.class_predictions(self.model.y_test_predict)
         test_df[prediction_column] = test_class_pred
 
@@ -704,10 +704,10 @@ class RobustnessDiagnosis(ThresholdTest):
         if self.model.train_ds.target_column in features_list:
             features_list.remove(self.model.train_ds.target_column)
 
-        train_df = self.model.train_ds.x.copy(deep=True)
+        train_df = self.model.train_ds.x.copy()
         train_y_true = self.model.train_ds.y
 
-        test_df = self.model.test_ds.x.copy(deep=True)
+        test_df = self.model.test_ds.x.copy()
         test_y_true = self.model.test_ds.y
 
         test_results = []
@@ -719,8 +719,8 @@ class RobustnessDiagnosis(ThresholdTest):
 
         # Iterate scaling factor for the standard deviation list
         for x_std_dev in x_std_dev_list:
-            temp_train_df = train_df.copy(deep=True)
-            temp_test_df = test_df.copy(deep=True)
+            temp_train_df = train_df.copy()
+            temp_test_df = test_df.copy()
 
             # Add noise to numeric features columns provided by user
             for feature in features_list: