si_lab04

from scipy.io import wavfile
import numpy as np
import pandas as pd
import os
import pandas as pd
import numpy as np
from sklearn.neighbors import KNeighborsClassifier as kNN
from sklearn.svm import SVC as SVM
from sklearn.metrics import confusion_matrix
from sklearn.model_selection import train_test_split
col_names = ['Gender', 'Married', 'Education', 'Self_Employed', 'Loan_Status']
col_values = ['Male', 'Yes', 'Graduate', 'Yes', 'Y']
pd.options.mode.chained_assignment = None
data = pd.read_excel('loan_data.xlsx');
for i in range(len(col_names)):
    mask = data[col_names[i]].values == col_values[i]
    data[col_names[i]][mask] = 0
    data[col_names[i]][~mask] = 1
columns = list(data.columns);
# mask = data ['Gender'].values == 'Female'
# data['Gender'][mask] = 1
# data['Gender'][~mask] = 0
cat_feature = pd.Categorical(data.Property_Area)
one_hot = pd.get_dummies(cat_feature)
data = pd.concat([data, one_hot], axis = 1)
data = data.drop(columns = ['Property_Area'])
features = data.columns
vals = data.values.astype(np.float64)
X = vals[:, :-1]
y = vals[:,-1]
models = [kNN(), SVM()]
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42)
for model in models:
    model.fit(X_train,y_train)
    y_pred = model.predict(X_test)
    print(confusion_matrix(y_test, y_pred))
from sklearn.tree import DecisionTreeClassifier as DT
from sklearn.tree import plot_tree
model = DT(max_depth=3)
model.fit(X_train, y_train)
y_pred = model.predict(X_test)
cm = confusion_matrix(y_test, y_pred)
print(cm)
from matplotlib import pyplot as plt
plt.figure(figsize=(20,10))
tree_vis = plot_tree(model,feature_names=
  data.columns[:-1],class_names=['N', 'Y'], fontsize = 20)
import numpy as np
from matplotlib import rcParams
from matplotlib import pyplot as plt
from sklearn.decomposition import PCA
example = np.random.randn(500,2)
example[:,1] *= 0.4
rot_matrix = np.array([[1/2**0.5, 1/2**0.5],
[1/2**0.5, -1/2**0.5]])
example = np.dot(example, rot_matrix)
example_PCAed = PCA(2).fit_transform(example)
rcParams['font.size'] = 32
rcParams['font.family'] = 'Times New Roman'
fig, ax = plt.subplots(1,2, figsize=(20, 10))
ax[0].scatter(example[:,0], example[:,1])
ax[1].scatter(example_PCAed[:,0], example_PCAed[:,1])
ax[0].set_xlim([-3, 3])
ax[0].set_ylim([-3, 3])
ax[1].set_xlim([-3, 3])
ax[1].set_ylim([-3, 3])
ax[0].set_xlabel('x')
ax[0].set_ylabel('y')
ax[1].set_xlabel('PC 1')
ax[1].set_ylabel('PC 2')
ax[0].set_title('Dane pierwotne')
ax[1].set_title('Dane po PCA')

pca_transform = PCA()
pca_transform.fit(X_train)
variances = pca_transform.explained_variance_ratio_
cumulated_variances = variances.cumsum()
plt.scatter(np.arange(variances.shape[0]),
cumulated_variances)
plt.yticks(np.arange(0, 1.1, 0.1))
PC_num = (cumulated_variances<0.95).sum()

X_pcaed = PCA(2).fit_transform(X_train)
fig, ax = plt.subplots(1,1)
females = y_train==1
ax.scatter(X_pcaed[females,0], X_pcaed[females,1], label =
'female')
ax.scatter(X_pcaed[~females,0], X_pcaed[~females,1], label
= 'males')
ax.legend()