import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
from logisticRegression.logisticRegression import LogisticRegression
from sklearn.datasets import load_breast_cancer
from sklearn.preprocessing import MinMaxScaler
from metrics import *
import math
scalar = MinMaxScaler()
data = load_breast_cancer()
X = pd.DataFrame(data['data'])
y = pd.Series(data['target'])
scalar.fit(X)
X = scalar.transform(X)
X = pd.DataFrame(X) # This scales data to the range 0-1 and is easier to train
LR = LogisticRegression()
LR.fit(X, y, n_iter=500, lr=5e-3)
y_hat = LR.predict(X)
print('Accuracy: ', accuracy(y_hat, y))
LR = LogisticRegression()
LR.fit_autograd(X, y, n_iter=400, lr=8e-3)
y_hat = LR.predict(X)
print('Accuracy: ', accuracy(y_hat, y))
for i in range(3):
xti = X.iloc[i * 190:min(570, (i + 1) * 190)]
yti = y[i * 190:min(570, (i + 1) * 190)]
xi1 = X.iloc[0:i * 190]
xi2 = X.iloc[min(570, (i + 1) * 190):570]
yi1 = y[0:i * 190]
yi2 = y[min(570, (i + 1) * 190):570]
xi = pd.concat([xi1, xi2])
yi = pd.concat([yi1, yi2])
xi.reset_index(drop=True, inplace=True)
yi.reset_index(drop=True, inplace=True)
xti.reset_index(drop=True, inplace=True)
yti.reset_index(drop=True, inplace=True)
LR = LogisticRegression()
LR.fit(xi, yi, n_iter=600, lr=7e-03)
y_hat = LR.predict(xti)
acc_curr = accuracy(y_hat, yti)
if (acc_curr > best_accuracy):
best_accuracy = acc_curr
best_LR = copy.deepcopy(LR)
print(f'Accuracy Fold {i+1}: ', acc_curr)
acc_ov += acc_curr
print("Overall Accuracy:", acc_ov / 3)
print("Decision boundary for features 0 and 1:")
LR = LogisticRegression()
LR.fit(X[[0, 1]], y, n_iter=500, lr=5e-3)
LR.plot(np.array(X[[0, 1]]), np.array(y))
from sklearn.preprocessing import MinMaxScaler
from sklearn.model_selection import KFold
scaler = MinMaxScaler()
np.random.seed(42)
from sklearn.datasets import load_breast_cancer
X = load_breast_cancer().data
y = load_breast_cancer().target
scalar = MinMaxScaler()
scalar.fit(X)
X = scalar.transform(X)
print('\n---------Unregularised Normal Fit----------')
LR = LogisticRegression(learningRate=0.1, maxIterations=1000)
LR.fit(X, y)
y_hat = LR.predict(X)
print(accuracy(y_hat, y))
print('\n---------Unregularised Autograd Fit----------')
LR = LogisticRegression(learningRate=0.1, maxIterations=1000)
LR.fit_autograd(X, y)
y_hat = LR.predict(X)
print(accuracy(y_hat, y))
print('\n---------3 KFold----------')
Kfolding = KFold(3, True, 1)
avgAccuracy = 0
for trainData, testData in Kfolding.split(X):
trainSetData, testSetData = X[trainData], X[testData]
ytrainData, ytestData = y[trainData], y[testData]