Spaces:
Runtime error
Runtime error
Update src/streamlit_app.py
Browse files- src/streamlit_app.py +57 -1
src/streamlit_app.py
CHANGED
|
@@ -3,7 +3,63 @@
|
|
| 3 |
import streamlit as st
|
| 4 |
import pandas as pd
|
| 5 |
# import matplotlib.pyplot as plt
|
| 6 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 7 |
from sklearn.metrics import r2_score
|
| 8 |
|
| 9 |
st.title("📈 Simple Linear Regression App")
|
|
|
|
| 3 |
import streamlit as st
|
| 4 |
import pandas as pd
|
| 5 |
# import matplotlib.pyplot as plt
|
| 6 |
+
|
| 7 |
+
class LinearRegression:
|
| 8 |
+
def __init__(self, learning_rate=0.01, epochs=1000):
|
| 9 |
+
self.learning_rate = learning_rate
|
| 10 |
+
self.epochs = epochs
|
| 11 |
+
self.w1 = 0.0
|
| 12 |
+
self.w2 = 0.0
|
| 13 |
+
self.b = 0.0
|
| 14 |
+
|
| 15 |
+
def predict(self, x1, x2):
|
| 16 |
+
return self.w1 * x1 + self.w2 * x2 + self.b
|
| 17 |
+
|
| 18 |
+
def compute_gradients(self, X1, X2, y):
|
| 19 |
+
n = len(y)
|
| 20 |
+
dw1, dw2, db = 0.0, 0.0, 0.0
|
| 21 |
+
|
| 22 |
+
for i in range(n):
|
| 23 |
+
y_pred = self.predict(X1[i], X2[i])
|
| 24 |
+
error = y[i] - y_pred
|
| 25 |
+
dw1 += -2 * X1[i] * error
|
| 26 |
+
dw2 += -2 * X2[i] * error
|
| 27 |
+
db += -2 * error
|
| 28 |
+
|
| 29 |
+
return dw1 / n, dw2 / n, db / n
|
| 30 |
+
|
| 31 |
+
def update(self, X1, X2, y):
|
| 32 |
+
dw1, dw2, db = self.compute_gradients(X1, X2, y)
|
| 33 |
+
self.w1 -= self.learning_rate * dw1
|
| 34 |
+
self.w2 -= self.learning_rate * dw2
|
| 35 |
+
self.b -= self.learning_rate * db
|
| 36 |
+
|
| 37 |
+
def mean_squared_error(self, y, y_pred):
|
| 38 |
+
n = len(y)
|
| 39 |
+
return sum((y[i] - y_pred[i]) ** 2 for i in range(n)) / n
|
| 40 |
+
|
| 41 |
+
def fit(self, X1, X2, y):
|
| 42 |
+
for epoch in range(self.epochs):
|
| 43 |
+
self.update(X1, X2, y)
|
| 44 |
+
|
| 45 |
+
if epoch % 200 == 0:
|
| 46 |
+
y_pred = [self.predict(X1[i], X2[i]) for i in range(len(y))]
|
| 47 |
+
mse = self.mean_squared_error(y, y_pred)
|
| 48 |
+
print(f"Epoch {epoch}: MSE = {mse:.4f}")
|
| 49 |
+
|
| 50 |
+
|
| 51 |
+
# -------------------------------
|
| 52 |
+
# TRAINING
|
| 53 |
+
# -------------------------------
|
| 54 |
+
X1 = [2, 4, 6, 8]
|
| 55 |
+
X2 = [2, 4, 6, 8]
|
| 56 |
+
y = [4, 8, 12, 16]
|
| 57 |
+
|
| 58 |
+
model = LinearRegression(learning_rate=0.01, epochs=1000)
|
| 59 |
+
model.fit(X1, X2, y)
|
| 60 |
+
|
| 61 |
+
print("Prediction before saving:", model.predict(2, 2))
|
| 62 |
+
|
| 63 |
from sklearn.metrics import r2_score
|
| 64 |
|
| 65 |
st.title("📈 Simple Linear Regression App")
|