Create app.py
Browse files
app.py
ADDED
@@ -0,0 +1,115 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
import gradio as gr
|
2 |
+
import joblib
|
3 |
+
import numpy as np
|
4 |
+
import pandas as pd
|
5 |
+
from huggingface_hub import hf_hub_download
|
6 |
+
from sklearn.preprocessing import StandardScaler, OneHotEncoder, LabelEncoder
|
7 |
+
|
8 |
+
# Load the trained model and scaler objects from file
|
9 |
+
|
10 |
+
|
11 |
+
REPO_ID = "Hemg/modelxxx" # hugging face repo ID
|
12 |
+
MoDEL_FILENAME = "studentpredict.joblib" # model file name
|
13 |
+
SCALER_FILENAME ="studentscaler.joblib" # scaler file name
|
14 |
+
|
15 |
+
model = joblib.load(hf_hub_download(repo_id=REPO_ID, filename=MoDEL_FILENAME))
|
16 |
+
|
17 |
+
scaler = joblib.load(hf_hub_download(repo_id=REPO_ID, filename=SCALER_FILENAME))
|
18 |
+
|
19 |
+
def encode_categorical_columns(df):
|
20 |
+
label_encoder = LabelEncoder()
|
21 |
+
ordinal_columns = df.select_dtypes(include=['object']).columns
|
22 |
+
|
23 |
+
for col in ordinal_columns:
|
24 |
+
df[col] = label_encoder.fit_transform(df[col])
|
25 |
+
|
26 |
+
nominal_columns = df.select_dtypes(include=['object']).columns.difference(ordinal_columns)
|
27 |
+
df = pd.get_dummies(df, columns=nominal_columns, drop_first=True)
|
28 |
+
|
29 |
+
return df
|
30 |
+
|
31 |
+
# Define the prediction function
|
32 |
+
def predict_performance(Location, College_Fee, GPA, Year, Course_Interested, College,
|
33 |
+
Faculty, Source, visited_college_for_inquiry_only, attended_event_name,
|
34 |
+
attended_any_events, Presenter, visited_with_parents):
|
35 |
+
|
36 |
+
# Debug print 1: Print all input values
|
37 |
+
print("\nInput Values:")
|
38 |
+
print(f"Location: {Location}")
|
39 |
+
print(f"College_Fee: {College_Fee}")
|
40 |
+
print(f"GPA: {GPA}")
|
41 |
+
print(f"Year: {Year}")
|
42 |
+
print(f"Course_Interested: {Course_Interested}")
|
43 |
+
print(f"Faculty: {Faculty}")
|
44 |
+
print(f"Source: {Source}")
|
45 |
+
print(f"Visited for inquiry: {visited_college_for_inquiry_only}")
|
46 |
+
print(f"Event name: {attended_event_name}")
|
47 |
+
print(f"Attended events: {attended_any_events}")
|
48 |
+
print(f"Presenter: {Presenter}")
|
49 |
+
print(f"Visited with parents: {visited_with_parents}")
|
50 |
+
|
51 |
+
input_data = [[Location, College_Fee, GPA, Year, Course_Interested, College,
|
52 |
+
Faculty, Source, visited_college_for_inquiry_only, attended_event_name,
|
53 |
+
attended_any_events, Presenter, visited_with_parents]]
|
54 |
+
|
55 |
+
feature_names = ["Location", "College_Fee", "GPA", "Year", "Course_Interested", "College",
|
56 |
+
"Faculty", "Source", "visited_college_for_inquiry_only", "attended_event_name",
|
57 |
+
"attended_any_events", "Presenter", "visited_with_parents"]
|
58 |
+
|
59 |
+
input_df = pd.DataFrame(input_data, columns=feature_names)
|
60 |
+
|
61 |
+
# Debug print 2: Show DataFrame before encoding
|
62 |
+
print("\nDataFrame before encoding:")
|
63 |
+
print(input_df)
|
64 |
+
|
65 |
+
df = encode_categorical_columns(input_df)
|
66 |
+
|
67 |
+
# Debug print 3: Show DataFrame after encoding
|
68 |
+
print("\nDataFrame after encoding:")
|
69 |
+
print(df)
|
70 |
+
|
71 |
+
scaled_input = scaler.transform(df)
|
72 |
+
|
73 |
+
# Debug print 4: Show scaled input
|
74 |
+
print("\nScaled input:")
|
75 |
+
print(scaled_input)
|
76 |
+
|
77 |
+
prediction = model.predict(scaled_input)[0]
|
78 |
+
|
79 |
+
# Debug print 5: Show prediction details
|
80 |
+
print("\nPrediction details:")
|
81 |
+
print(f"Raw prediction: {prediction}")
|
82 |
+
prediction_probability = 1 / (1 + np.exp(-prediction))
|
83 |
+
print(f"Probability: {prediction_probability}")
|
84 |
+
prediction_percentage = prediction_probability * 100
|
85 |
+
print(f"Percentage: {prediction_percentage}")
|
86 |
+
|
87 |
+
return f"Chance of Admission: {prediction_percentage:.1f}%"
|
88 |
+
|
89 |
+
iface = gr.Interface(
|
90 |
+
fn=predict_performance,
|
91 |
+
inputs=[
|
92 |
+
gr.Radio(["Kathmandu", "Bhaktapur", "Lalitpur", "Kritipur"], label="Location"),
|
93 |
+
gr.Slider(minimum=1000000, maximum=1700000, label="College Fee"),
|
94 |
+
gr.Slider(minimum=2, maximum=3, label="GPA"), # Fixed GPA input
|
95 |
+
gr.Slider(minimum=2019, maximum=2025, step=1, label="Year"),
|
96 |
+
gr.Radio(["MSc IT & Applied Security", "BSc (Hons) Computing", "BSc (Hons) Computing with Artificial Intelligence",
|
97 |
+
"BSc (Hons) Computer Networking & IT Security", "BSc (Hons) Multimedia Technologies", "MBA",
|
98 |
+
"BA (Hons) Accounting & Finance", "BA (Hons) Business Administration"], label="Course_Interested"),
|
99 |
+
gr.Radio(["Science", "Management", "Humanities"], label="Faculty"),
|
100 |
+
gr.Radio(["Event", "Facebook", "Instagram", "Offline", "Recommendation"], label="Source"),
|
101 |
+
gr.Radio(["Yes", "No"], label="visited_college_for_inquery_only"),
|
102 |
+
gr.Radio(["New Year", "Dashain", "Orientation", "Fresher's Party", "Holi Festival", "Welcome Ceremony"],
|
103 |
+
label="attended_event_name"), # Changed from Slider to Radio
|
104 |
+
gr.Radio(["Yes", "No"], label="attended_any_event"),
|
105 |
+
gr.Radio(["Ram", "Gita", "Manish", "Shyam", "Raj", "Hari", "Rina", "Shree"], label="Presenter"),
|
106 |
+
gr.Radio(["Yes", "No"], label="visited_with_parents")
|
107 |
+
],
|
108 |
+
outputs="text",
|
109 |
+
title="chances of student admission",
|
110 |
+
description="chances of student admission"
|
111 |
+
)
|
112 |
+
|
113 |
+
# Run the app
|
114 |
+
if __name__ == "__main__":
|
115 |
+
iface.launch(share=True)
|