import numpy as np
import matplotlib.pyplot as plt
from getDataset import getDataSet
from sklearn.linear_model import LogisticRegression
# Starting codes
# Fill in the codes between "%PLACEHOLDER#start" and "PLACEHOLDER#end"
# step 1: generate dataset that includes both positive and negative samples,
# where each sample is described with two features.
# 250 samples in total.
[X, y] = getDataSet() # note that y contains only 1s and 0s,
# create figure for all charts to be placed on so can be viewed together
fig = plt.figure()
def func_DisplayData(dataSamplesX, dataSamplesY, chartNum, titleMessage):
print(dataSamplesY == 0)
idx1 = (dataSamplesY == 0).nonzero() # object indices for the 1st class
idx2 = (dataSamplesY == 1).nonzero()
ax = fig.add_subplot(1, 3, chartNum)
fig.set_size_inches(15, 5, forward=True)
# no more variables are needed
plt.plot(dataSamplesX[idx1, 0], dataSamplesX[idx1, 1], 'r*')
plt.plot(dataSamplesX[idx2, 0], dataSamplesX[idx2, 1], 'b*')
# axis tight
ax.set_xlabel('x_1')
ax.set_ylabel('x_2')
# tp True Positive
tn, fp, fn, tp = confMatrix.ravel()
print('True Positive: ' + str(tp))
print('False Positive: ' + str(fp))
print('True Negative: ' + str(tn))
print('False Negative: ' + str(fn))
recall = tp/(tp + fn)
precision = tp/(tp + fp)
accuracy = (tp + tn)/(tp+fp+tn+fn)
print('Accuracy: {}'.format(accuracy))
print('Precision: {}'.format(precision))
print('Recall: {}'.format(recall))
# Creating a random sample set
X,y = getDataSet()
# Create figure for all charts to be placed on so can be viewed together
fig = plt.figure()
# Plotting all samples
func_DisplayData(X, y, 1, 'All samples')
# Splitting the sample set into training and testing samples
trainX,testX,trainY,testY = train_test_split(X,y,test_size = 0.3)
# Plot the samples you have pickup for training, check to confirm that both negative
# and positive samples are included.
func_DisplayData(trainX, trainY, 2, 'training samples')
func_DisplayData(testX, testY, 3, 'testing samples')
# Show all charts
plt.show()
# Use sklearn logistic regression model
clf = LogisticRegression()