ajey091

sampleddata = dist.rvs(200)
plt.hist(sampleddata,bins=20)
plt.xlabel('Sampled men height in cm')
plt.ylabel('PDF')
plt.show()

ajey091

x,y = eval_pdf(dist)
plt.plot(x,y)
plt.xlabel('Men height in cm')
plt.ylabel('PDF')
plt.show()

ajey091

# We will create a random variable object using scipy.stats
mu = 178
sigma = 7.7
dist = scipy.stats.norm(mu,sigma) # Creates a random variable object

def eval_pdf(rv):
    mu = rv.mean()
    sigma = rv.std()
    xs = np.linspace(mu-4*sigma,mu+4*sigma,100) # defining x values as a range around the mean
    ys = rv.pdf(xs)

ajey091

import numpy as np 
import scipy.stats
import matplotlib.pyplot as plt

ajey091

ajey091

rf_acc = []
rf_f1 = []
for rs in range(100):
    X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.25, random_state=rs)
    y_train = np.ravel(y_train)
    y_test = np.ravel(y_test)
    rfclf = RandomForestClassifier(max_depth=50, random_state=0,n_estimators=100)
    rfclf.fit(X_train, y_train)
    y_pred_rf = rfclf.predict(X_test)
    rf_acc.append(accuracy_score(y_test, y_pred_rf))

ajey091

# Random Forest classifier
from sklearn.ensemble import RandomForestClassifier
rfclf = RandomForestClassifier(max_depth=50, random_state=0,n_estimators=1000)
rfclf.fit(X_train, y_train)
y_pred_rf = rfclf.predict(X_test)

rf_probs = rfclf.predict_proba(X_test)
rf_probs = rf_probs[:, 1]

print('Random Forest: Accuracy=%.3f' % (accuracy_score(y_test, y_pred_rf)))

ajey091

# Training the data and predicting 
# Logistic regression
from sklearn.linear_model import LogisticRegression
# fit a model
lrclf = LogisticRegression(solver='lbfgs',max_iter = 10000)
lrclf.fit(X_train, y_train)
y_pred_lr = lrclf.predict(X_test)

lr_probs = lrclf.predict_proba(X_test)
lr_probs = lr_probs[:, 1]

ajey091

# Loading a bunch of sklearn modules:
from sklearn.metrics import accuracy_score
from sklearn.dummy import DummyClassifier
from sklearn.metrics import confusion_matrix
from sklearn.metrics import f1_score
from sklearn.metrics import roc_curve
from sklearn.metrics import roc_auc_score

dummy_clf = DummyClassifier(strategy="most_frequent")
dummy_clf.fit(X_train, y_train)

ajey091

print (y['binnedgrade'].value_counts()) # we clearly have imbalanced datasets.