Spaces:
Runtime error
Runtime error
app.py
Browse files
app.py
ADDED
@@ -0,0 +1,112 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
import numpy as np
|
2 |
+
import pandas as pd
|
3 |
+
import matplotlib.pyplot as plt
|
4 |
+
import seaborn as sns
|
5 |
+
import plotly.express as px
|
6 |
+
from sklearn.tree import DecisionTreeClassifier
|
7 |
+
from sklearn.ensemble import RandomForestClassifier
|
8 |
+
from sklearn.model_selection import RandomizedSearchCV
|
9 |
+
|
10 |
+
df = pd.read_csv("credit_risk_dataset.csv")
|
11 |
+
|
12 |
+
df = df.dropna()
|
13 |
+
|
14 |
+
df.columns
|
15 |
+
|
16 |
+
X =df.drop("loan_status", axis = 1)
|
17 |
+
y = df['loan_status']
|
18 |
+
|
19 |
+
categorical_features = ["person_home_ownership", "loan_intent", "loan_grade", "cb_person_default_on_file"]
|
20 |
+
X = pd.get_dummies(X, categorical_features)
|
21 |
+
X.columns
|
22 |
+
|
23 |
+
|
24 |
+
|
25 |
+
from sklearn.model_selection import train_test_split
|
26 |
+
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size = 0.2)
|
27 |
+
|
28 |
+
X_train.head()
|
29 |
+
|
30 |
+
from sklearn.preprocessing import StandardScaler
|
31 |
+
scaler_normal = StandardScaler()
|
32 |
+
def scaler(data, runtime = False):
|
33 |
+
normal_col = ['person_income','person_age','person_emp_length', 'loan_amnt','loan_int_rate','cb_person_cred_hist_length','loan_percent_income']
|
34 |
+
if(runtime == False):
|
35 |
+
data.loc[:,normal_col] = scaler_normal.fit_transform(data.loc[:,normal_col])
|
36 |
+
else:
|
37 |
+
data.loc[:,normal_col] = scaler_normal.transform(data.loc[:,normal_col])
|
38 |
+
return data
|
39 |
+
|
40 |
+
X_train = scaler(X_train)
|
41 |
+
X_test = scaler(X_test, True)
|
42 |
+
|
43 |
+
model = RandomForestClassifier(max_depth = 5)
|
44 |
+
model.fit(X_train, y_train)
|
45 |
+
|
46 |
+
y_predict = model.predict(X_test)
|
47 |
+
|
48 |
+
features = {
|
49 |
+
"person_home_ownership": ['MORTGAGE', 'OTHER','OWN', 'RENT',],
|
50 |
+
"loan_intent": ['DEBTCONSOLIDATION', 'EDUCATION', 'HOMEIMPROVEMENT', 'MEDICAL', 'PERSONAL', 'VENTURE'],
|
51 |
+
"loan_grade": ['A','B', 'C', 'D', 'E','F', 'G'],
|
52 |
+
"cb_person_default_on_file": ['N', 'Y']
|
53 |
+
}
|
54 |
+
def preprocess(model_input):
|
55 |
+
for feature in features:
|
56 |
+
for option in features[feature]:
|
57 |
+
selection = model_input[feature]
|
58 |
+
if option is selection:
|
59 |
+
model_input[f'{feature}_{option}'] = 1
|
60 |
+
else:
|
61 |
+
model_input[f'{feature}_{option}'] = 0
|
62 |
+
|
63 |
+
model_input.drop([_ for _ in features], inplace = True, axis = 1)
|
64 |
+
return model_input
|
65 |
+
|
66 |
+
|
67 |
+
|
68 |
+
def credit_run(person_age, person_income, person_home_ownership,
|
69 |
+
person_emp_length, loan_intent, loan_grade, loan_amnt,
|
70 |
+
loan_int_rate, cb_person_default_on_file, cb_person_cred_hist_length):
|
71 |
+
model_input = preprocess(
|
72 |
+
pd.DataFrame( { 'person_age': person_age,
|
73 |
+
'person_income': person_income,
|
74 |
+
'person_home_ownership': person_home_ownership,
|
75 |
+
'person_emp_length': person_emp_length,
|
76 |
+
'loan_intent': loan_intent,
|
77 |
+
'loan_grade': loan_grade,
|
78 |
+
'loan_amnt': loan_amnt,
|
79 |
+
'loan_int_rate': loan_int_rate,
|
80 |
+
'loan_percent_income': loan_amnt / person_income,
|
81 |
+
'cb_person_default_on_file': cb_person_default_on_file,
|
82 |
+
'cb_person_cred_hist_length': cb_person_cred_hist_length
|
83 |
+
}, index = [0]
|
84 |
+
))
|
85 |
+
out = model.predict(model_input)
|
86 |
+
return "High risk of defaulting" if out[0] == 1 else "Low risk of defaulting"
|
87 |
+
|
88 |
+
|
89 |
+
import gradio as gr
|
90 |
+
demo = gr.Interface(
|
91 |
+
fn = credit_run,
|
92 |
+
inputs = [
|
93 |
+
gr.Slider(label="Person_Age(In Years)", minimum=18, maximum=90, step=1),
|
94 |
+
gr.Number(label="person_income(per month)"),
|
95 |
+
gr.Radio(['MORTGAGE', 'OTHER','OWN', 'RENT']),
|
96 |
+
gr.Slider(label="person_emp_length(In Years)", minimum=0, maximum=60, step=1),
|
97 |
+
gr.Radio(['DEBTCONSOLIDATION', 'EDUCATION', 'HOMEIMPROVEMENT', 'MEDICAL', 'PERSONAL', 'VENTURE']),
|
98 |
+
gr.Radio(['A','B', 'C', 'D', 'E','F', 'G']),
|
99 |
+
gr.Number(label="loan_amnt"),
|
100 |
+
gr.Number(label="loan_int_rate"),
|
101 |
+
gr.Radio(['0', '1']),
|
102 |
+
gr.Number(label="cb_person_cred_hist_length"),
|
103 |
+
],
|
104 |
+
outputs = "text",
|
105 |
+
title = "Credit Risk Predictor",
|
106 |
+
examples = [[23,15000,'RENT',2,'EDUCATION','A',300000,8.9,'0',6],
|
107 |
+
[34,50000,'OWN',1,'MEDICAL','B',1000000,10.65,'0',3],
|
108 |
+
[42,30000,"MORTGAGE",12,'HOMEIMPROVEMENT','C',800000,7.9,'1',8],
|
109 |
+
[42,20000,"MORTGAGE",10,'PERSONAL','F',100000,15.25,'1',5]]
|
110 |
+
)
|
111 |
+
demo.launch(share = True,debug=True)
|
112 |
+
|