Spaces:
Sleeping
Sleeping
Create voting_classifier_viz.py
Browse files- voting_classifier_viz.py +74 -0
voting_classifier_viz.py
ADDED
|
@@ -0,0 +1,74 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
import streamlit as st
|
| 2 |
+
import matplotlib.pyplot as plt
|
| 3 |
+
from sthelper import StHelper
|
| 4 |
+
import data_helper
|
| 5 |
+
|
| 6 |
+
# Import all datasets
|
| 7 |
+
concentric, linear, outlier, spiral, ushape, xor = data_helper.load_dataset()
|
| 8 |
+
|
| 9 |
+
# Configure matplotlib styling
|
| 10 |
+
plt.style.use('seaborn-v0_8-bright')
|
| 11 |
+
|
| 12 |
+
# Dataset selection dropdown
|
| 13 |
+
st.sidebar.markdown("# Voting Classifier")
|
| 14 |
+
dataset = st.sidebar.selectbox(
|
| 15 |
+
"Dataset",
|
| 16 |
+
("U-Shaped", "Linearly Separable", "Outlier", "Two Spirals", "Concentric Circles", "XOR")
|
| 17 |
+
)
|
| 18 |
+
|
| 19 |
+
# Estimator multi-select
|
| 20 |
+
estimators = st.sidebar.multiselect(
|
| 21 |
+
'Estimators',
|
| 22 |
+
[
|
| 23 |
+
'KNN',
|
| 24 |
+
'Logistic Regression',
|
| 25 |
+
'Gaussian Naive Bayes',
|
| 26 |
+
'SVM',
|
| 27 |
+
'Random Forest'
|
| 28 |
+
]
|
| 29 |
+
)
|
| 30 |
+
|
| 31 |
+
# Voting type radio button
|
| 32 |
+
voting_type = st.sidebar.radio(
|
| 33 |
+
"Voting Type",
|
| 34 |
+
(
|
| 35 |
+
'hard',
|
| 36 |
+
'soft',
|
| 37 |
+
)
|
| 38 |
+
)
|
| 39 |
+
|
| 40 |
+
st.header(dataset)
|
| 41 |
+
fig, ax = plt.subplots()
|
| 42 |
+
|
| 43 |
+
# Plot initial graph
|
| 44 |
+
df = data_helper.load_initial_graph(dataset, ax)
|
| 45 |
+
orig = st.pyplot(fig)
|
| 46 |
+
|
| 47 |
+
# Extract X and Y
|
| 48 |
+
X = df.iloc[:, :2].values
|
| 49 |
+
y = df.iloc[:, -1].values
|
| 50 |
+
|
| 51 |
+
# Create sthelper object
|
| 52 |
+
sthelper = StHelper(X, y)
|
| 53 |
+
|
| 54 |
+
# On button click
|
| 55 |
+
if st.sidebar.button("Run Algorithm"):
|
| 56 |
+
algos = sthelper.create_base_estimators(estimators, voting_type)
|
| 57 |
+
voting_clf, voting_clf_accuracy = sthelper.train_voting_classifier(algos, voting_type)
|
| 58 |
+
sthelper.draw_main_graph(voting_clf, ax)
|
| 59 |
+
orig.pyplot(fig)
|
| 60 |
+
figs = sthelper.plot_other_graphs(algos)
|
| 61 |
+
|
| 62 |
+
# Plot accuracies
|
| 63 |
+
st.sidebar.header("Classification Metrics")
|
| 64 |
+
st.sidebar.text("Voting Classifier accuracy: " + str(voting_clf_accuracy))
|
| 65 |
+
accuracies = sthelper.calculate_base_model_accuracy(algos)
|
| 66 |
+
for i in range(len(accuracies)):
|
| 67 |
+
st.sidebar.text("Accuracy for Model " + str(i + 1) + " - " + str(accuracies[i]))
|
| 68 |
+
|
| 69 |
+
counter = 0
|
| 70 |
+
for i in st.columns(len(figs)):
|
| 71 |
+
with i:
|
| 72 |
+
st.pyplot(figs[counter])
|
| 73 |
+
st.text(counter)
|
| 74 |
+
counter += 1
|