Mushroom_Classification/app.py at master · TargetTactician/Mushroom_Classification · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
import streamlit as st
import pandas as pd
import pickle
import os

# Optional: set working directory (update if needed)
os.chdir('/home/parthi/Git/Data Science/Machine Learning/Mushroom_Classification/')

# Load model + encoders
with open('model_with_encoders.pkl', 'rb') as f:
    bundle = pickle.load(f)

model = bundle['model']
label_encoders = bundle['label_encoders']
feature_columns = bundle['feature_columns']

# Streamlit Title
st.title("🍄 Mushroom Classification App")
st.write("Predict whether a mushroom is edible or poisonous based on its characteristics.")

# ---- FILE UPLOAD OR MANUAL INPUT ---- #
st.sidebar.header("Choose Input Method")
input_method = st.sidebar.radio("Select input method:", ["Manual Entry", "Upload CSV"])

# ---------- INPUT METHOD: MANUAL ---------- #
if input_method == "Manual Entry":
    st.subheader("Manual Input")

    # Use only the values seen during training
    cap_shape = st.selectbox("Cap Shape", label_encoders['cap-shape'].classes_)
    cap_surface = st.selectbox("Cap Surface", label_encoders['cap-surface'].classes_)
    cap_color = st.selectbox("Cap Color", label_encoders['cap-color'].classes_)
    bruises = st.selectbox("Bruises", label_encoders['bruises'].classes_)
    odor = st.selectbox("Odor", label_encoders['odor'].classes_)
    gill_attachment = st.selectbox("Gill Attachment", label_encoders['gill-attachment'].classes_)
    gill_spacing = st.selectbox("Gill Spacing", label_encoders['gill-spacing'].classes_)
    gill_size = st.selectbox("Gill Size", label_encoders['gill-size'].classes_)
    gill_color = st.selectbox("Gill Color", label_encoders['gill-color'].classes_)
    stalk_shape = st.selectbox("Stalk Shape", label_encoders['stalk-shape'].classes_)
    stalk_root = st.selectbox("Stalk Root", label_encoders['stalk-root'].classes_)
    stalk_surface_above_ring = st.selectbox("Stalk Surface Above Ring", label_encoders['stalk-surface-above-ring'].classes_)
    stalk_surface_below_ring = st.selectbox("Stalk Surface Below Ring", label_encoders['stalk-surface-below-ring'].classes_)
    stalk_color_above_ring = st.selectbox("Stalk Color Above Ring", label_encoders['stalk-color-above-ring'].classes_)
    stalk_color_below_ring = st.selectbox("Stalk Color Below Ring", label_encoders['stalk-color-below-ring'].classes_)
    veil_type = st.selectbox("Veil Type", label_encoders['veil-type'].classes_)
    veil_color = st.selectbox("Veil Color", label_encoders['veil-color'].classes_)
    ring_number = st.selectbox("Ring Number", label_encoders['ring-number'].classes_)
    ring_type = st.selectbox("Ring Type", label_encoders['ring-type'].classes_)
    spore_print_color = st.selectbox("Spore Print Color", label_encoders['spore-print-color'].classes_)
    population = st.selectbox("Population", label_encoders['population'].classes_)
    habitat = st.selectbox("Habitat", label_encoders['habitat'].classes_)

    # Build DataFrame
    input_data = [[
        cap_shape, cap_surface, cap_color, bruises, odor,
        gill_attachment, gill_spacing, gill_size, gill_color,
        stalk_shape, stalk_root, stalk_surface_above_ring,
        stalk_surface_below_ring, stalk_color_above_ring,
        stalk_color_below_ring, veil_type, veil_color, ring_number,
        ring_type, spore_print_color, population, habitat
    ]]
    input_df = pd.DataFrame(input_data, columns=feature_columns)

# ---------- INPUT METHOD: FILE UPLOAD ---------- #
elif input_method == "Upload CSV":
    st.subheader("Upload a CSV File")
    uploaded_file = st.file_uploader("Upload a CSV with same feature columns", type=['csv'])

    if uploaded_file is not None:
        input_df = pd.read_csv(uploaded_file)
        st.write("📄 Uploaded Data Preview:")
        st.write(input_df.head())

        # Validate columns
        if set(input_df.columns) != set(feature_columns):
            st.error("⚠️ Uploaded file must contain exactly these columns:")
            st.code(", ".join(feature_columns))
            st.stop()
    else:
        input_df = None

# ---------- PREDICT BUTTON (COMMON) ---------- #
if st.button("🔍 Predict"):
    if input_df is None:
        st.warning("Please upload a file or complete the manual inputs.")
        st.stop()

    # Encode using stored LabelEncoders
    try:
        for col in input_df.columns:
            if col in label_encoders:
                input_df[col] = label_encoders[col].transform(input_df[col])
    except Exception as e:
        st.error(f"Encoding error: {e}")
        st.stop()

    # Predict
    predictions = model.predict(input_df)
    result_labels = ['🟢 Edible' if p == 0 else '🔴 Poisonous' for p in predictions]

    input_df['Prediction'] = result_labels

    st.success("🎯 Prediction Results")
    st.dataframe(input_df[['Prediction']])

    # Optional download
    csv = input_df.to_csv(index=False)
    st.download_button("📥 Download Result CSV", csv, file_name="mushroom_predictions.csv")

    # Count predictions
    num_predictions = len(predictions)
    num_edible = sum(pred == 0 for pred in predictions)
    num_poisonous = sum(pred == 1 for pred in predictions)

    # Percentage
    pct_edible = round((num_edible / num_predictions) * 100, 2)
    pct_poisonous = round((num_poisonous / num_predictions) * 100, 2)

    # Show metrics
    st.subheader("📊 Prediction Summary")
    col1, col2, col3 = st.columns(3)

    col1.metric("Total Samples", num_predictions)
    col2.metric("🟢 Edible", f"{num_edible} ({pct_edible}%)")
    col3.metric("🔴 Poisonous", f"{num_poisonous} ({pct_poisonous}%)")