Duplicate from SoooSlooow/CreditCardsApp

Co-authored-by: Andrey Ovsyannikov <[email protected]>

.gitattributes

@@ -0,0 +1,33 @@
+*.7z filter=lfs diff=lfs merge=lfs -text
+*.arrow filter=lfs diff=lfs merge=lfs -text
+*.bin filter=lfs diff=lfs merge=lfs -text
+*.bz2 filter=lfs diff=lfs merge=lfs -text
+*.ftz filter=lfs diff=lfs merge=lfs -text
+*.gz filter=lfs diff=lfs merge=lfs -text
+*.h5 filter=lfs diff=lfs merge=lfs -text
+*.joblib filter=lfs diff=lfs merge=lfs -text
+*.lfs.* filter=lfs diff=lfs merge=lfs -text
+*.model filter=lfs diff=lfs merge=lfs -text
+*.msgpack filter=lfs diff=lfs merge=lfs -text
+*.npy filter=lfs diff=lfs merge=lfs -text
+*.npz filter=lfs diff=lfs merge=lfs -text
+*.onnx filter=lfs diff=lfs merge=lfs -text
+*.ot filter=lfs diff=lfs merge=lfs -text
+*.parquet filter=lfs diff=lfs merge=lfs -text
+*.pickle filter=lfs diff=lfs merge=lfs -text
+*.pkl filter=lfs diff=lfs merge=lfs -text
+*.pb filter=lfs diff=lfs merge=lfs -text
+*.pt filter=lfs diff=lfs merge=lfs -text
+*.pth filter=lfs diff=lfs merge=lfs -text
+*.rar filter=lfs diff=lfs merge=lfs -text
+saved_model/**/* filter=lfs diff=lfs merge=lfs -text
+*.tar.* filter=lfs diff=lfs merge=lfs -text
+*.tflite filter=lfs diff=lfs merge=lfs -text
+*.tgz filter=lfs diff=lfs merge=lfs -text
+*.wasm filter=lfs diff=lfs merge=lfs -text
+*.xz filter=lfs diff=lfs merge=lfs -text
+*.zip filter=lfs diff=lfs merge=lfs -text
+*.zst filter=lfs diff=lfs merge=lfs -text
+*tfevents* filter=lfs diff=lfs merge=lfs -text
+models/final_model.pkl filter=lfs diff=lfs merge=lfs -text
+models/other/unique_column_values.pkl filter=lfs diff=lfs merge=lfs -text

README.md

@@ -0,0 +1,13 @@
+---
+title: CreditCardsApp
+emoji: 💻
+colorFrom: gray
+colorTo: red
+sdk: gradio
+sdk_version: 3.3.1
+app_file: app.py
+pinned: false
+duplicated_from: SoooSlooow/CreditCardsApp
+---
+
+Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference

app.py

@@ -0,0 +1,167 @@
+import subprocess
+import random
+from typing import Any
+
+import gradio as gr
+import joblib
+import numpy as np
+import pandas as pd
+
+OUTPUT_DATA_PATH = "data/processed/app_dataset.csv"
+PREDICTIONS_PATH = "models/predictions/app_predictions.csv"
+UNIQUE_VALUES_PATH = "models/other/unique_column_values.pkl"
+MODEL_PATH = "models/final_model.pkl"
+
+
+def predict(*args: tuple) -> Any:
+    app_df = pd.DataFrame(data=[args], columns=columns, index=[0])
+    app_df.to_csv(OUTPUT_DATA_PATH, index=False)
+    model = joblib.load(MODEL_PATH)
+    predictions = model.predict_proba(app_df)
+    print(predictions)
+    if predictions[0][0] < 0.99:
+        message = "Client is considered bad. Issuance of credit is not recommended."
+    else:
+        message = "Client is considered good. Issuance of credit is allowed."
+    return round(predictions[0][0], 3), message
+
+
+columns = (
+    "YEARS_BIRTH",
+    "CODE_GENDER",
+    "AMT_INCOME_TOTAL",
+    "NAME_INCOME_TYPE",
+    "YEARS_EMPLOYED",
+    "OCCUPATION_TYPE",
+    "NAME_EDUCATION_TYPE",
+    "CNT_FAM_MEMBERS",
+    "CNT_CHILDREN",
+    "NAME_FAMILY_STATUS",
+    "FLAG_OWN_CAR",
+    "FLAG_OWN_REALTY",
+    "NAME_HOUSING_TYPE",
+    "FLAG_PHONE",
+    "FLAG_WORK_PHONE",
+    "FLAG_EMAIL",
+)
+unique_values = joblib.load(UNIQUE_VALUES_PATH)
+
+with gr.Blocks() as demo:
+    with gr.Row():
+        with gr.Column():
+            age = gr.Slider(label="Age", minimum=18, maximum=90, step=1, randomize=True)
+            sex = gr.Dropdown(
+                label="Sex",
+                choices=unique_values["CODE_GENDER"],
+                value=lambda: random.choice(unique_values["CODE_GENDER"]),
+            )
+            annual_income = gr.Slider(
+                label="Annual income",
+                minimum=0,
+                maximum=1000000,
+                step=10000,
+                randomize=True,
+            )
+            income_type = gr.Dropdown(
+                label="Income type",
+                choices=unique_values["NAME_INCOME_TYPE"],
+                value=lambda: random.choice(unique_values["NAME_INCOME_TYPE"]),
+            )
+            work_experience = gr.Slider(
+                label="Work experience at current position",
+                minimum=0,
+                maximum=75,
+                step=1,
+                randomize=True,
+            )
+            occupation_type = gr.Dropdown(
+                label="Occupation type",
+                choices=unique_values["OCCUPATION_TYPE"],
+                value=lambda: random.choice(unique_values["OCCUPATION_TYPE"]),
+            )
+            education_type = gr.Dropdown(
+                label="Education type",
+                choices=unique_values["NAME_EDUCATION_TYPE"],
+                value=lambda: random.choice(unique_values["NAME_EDUCATION_TYPE"]),
+            )
+            amount_of_family_members = gr.Slider(
+                label="Amount of family members",
+                minimum=0,
+                maximum=12,
+                step=1,
+                randomize=True,
+            )
+            amount_of_children = gr.Slider(
+                label="Amount of children",
+                minimum=0,
+                maximum=10,
+                step=1,
+                randomize=True,
+            )
+
+        with gr.Column():
+            family_status = gr.Dropdown(
+                label="Family status",
+                choices=unique_values["NAME_FAMILY_STATUS"],
+                value=lambda: random.choice(unique_values["NAME_FAMILY_STATUS"]),
+            )
+            flag_own_car = gr.Dropdown(
+                label="Having a car",
+                choices=unique_values["FLAG_OWN_REALTY"],
+                value=lambda: random.choice(unique_values["FLAG_OWN_REALTY"]),
+            )
+            flag_own_realty = gr.Dropdown(
+                label="Having a realty",
+                choices=unique_values["FLAG_OWN_REALTY"],
+                value=lambda: random.choice(unique_values["FLAG_OWN_REALTY"]),
+            )
+            housing_type = gr.Dropdown(
+                label="Housing type",
+                choices=unique_values["NAME_HOUSING_TYPE"],
+                value=lambda: random.choice(unique_values["NAME_HOUSING_TYPE"]),
+            )
+            flag_phone = gr.Dropdown(
+                label="Having a phone",
+                choices=unique_values["FLAG_PHONE"],
+                value=lambda: random.choice(unique_values["FLAG_PHONE"]),
+            )
+            flag_work_phone = gr.Dropdown(
+                label="Having a work phone",
+                choices=unique_values["FLAG_WORK_PHONE"],
+                value=lambda: random.choice(unique_values["FLAG_WORK_PHONE"]),
+            )
+            flag_email = gr.Dropdown(
+                label="Having an email",
+                choices=unique_values["FLAG_EMAIL"],
+                value=lambda: random.choice(unique_values["FLAG_EMAIL"]),
+            )
+
+        with gr.Column():
+            label_1 = gr.Label(label="Client rating")
+            label_2 = gr.Textbox(label="Client verdict (client is considered bad if client rating < 0.99)")
+            with gr.Row():
+                predict_btn = gr.Button(value="Predict")
+            predict_btn.click(
+                predict,
+                inputs=[
+                    age,
+                    sex,
+                    annual_income,
+                    income_type,
+                    work_experience,
+                    occupation_type,
+                    education_type,
+                    amount_of_family_members,
+                    amount_of_children,
+                    family_status,
+                    flag_own_car,
+                    flag_own_realty,
+                    housing_type,
+                    flag_phone,
+                    flag_work_phone,
+                    flag_email,
+                ],
+                outputs=[label_1, label_2],
+            )
+
+demo.launch()

data/processed/app_dataset.csv

@@ -0,0 +1,2 @@
+YEARS_BIRTH,CODE_GENDER,AMT_INCOME_TOTAL,NAME_INCOME_TYPE,YEARS_EMPLOYED,OCCUPATION_TYPE,NAME_EDUCATION_TYPE,CNT_FAM_MEMBERS,CNT_CHILDREN,NAME_FAMILY_STATUS,FLAG_OWN_CAR,FLAG_OWN_REALTY,NAME_HOUSING_TYPE,FLAG_PHONE,FLAG_WORK_PHONE,FLAG_EMAIL
+51,F,1540000,Pensioner,59,High skill tech staff,Academic degree,6,7,Separated,Yes,Yes,Rented apartment,Yes,Yes,No

models/final_model.pkl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:12919987df9acc9b14a30e37dcb292abf3a8d5586112c41e3384dabe006b7668
+size 1137472

models/other/unique_column_values.pkl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c662605d2104301585c83e2a6a9f7f3ceeae44bea0f08858618a8b3cbd11b523
+size 992

models/predictions/app_predictions.csv

@@ -0,0 +1,2 @@
+proba_0,proba_1,label
+0.9597238767799152,0.04027612322008478,1.0

src/models/make_predictions.py

@@ -0,0 +1,41 @@
+import click
+import joblib
+import numpy as np
+import pandas as pd
+
+
+@click.command()
+@click.argument("input_data_path", type=click.Path(exists=True))
+@click.argument("input_model_path", type=click.Path(exists=True))
+@click.argument("output_predictions_path", type=click.Path())
+def make_predictions(
+    input_data_path: str, input_model_path: str, output_predictions_path: str
+) -> None:
+    """
+    Предсказывает значения меток в входных данных, используя подаваемую на вход модель.
+    Предсказания записываются в csv-файл с тремя столбцами. В первые два столбца записываются вероятности
+    отнесения объекта к классу 0 и 1 соответственно, в третий - предсказываемая метка объекта на основе
+    выбранного порога вероятности.
+    :param input_data_path: путь к данным
+    :param input_model_path: путь к обученной модели
+    :param output_predictions_path: путь к файлу с получаемыми предсказаниями
+    """
+    df = pd.read_csv(input_data_path)
+    X = df.drop(["BAD_CLIENT"], axis=1, errors="ignore")
+
+    model = joblib.load(input_model_path)
+    probas = model.predict_proba(X)
+    labels = (probas[:, 1] > 0.01).astype(int)
+    predictions = pd.DataFrame(
+        data=np.column_stack([probas, labels]), columns=["proba_0", "proba_1", "label"]
+    )
+
+    predictions.to_csv(output_predictions_path, index=False)
+
+
+if __name__ == "__main__":
+    make_predictions()
+
+"""
+python -m src.models.make_predictions processed/processed/test_dataset.csv models/final_model.pkl reports/predictions.csv
+"""