statsmodels

Question 1

creates a new pd.df with a new column (equal in length to x1), which consists only of 1s to account for the constant beta_0

Answer 1

x = sm.add_constant(x1)

Question 2

Fit the model according to the OLS with a dependent variable y and an independent x

Answer 2

results = sm.OLS(y,x).fit()

Question 3

display model summary

Answer 3

results.summary()

Question 4

make prediction

Answer 4

predictions = results.predict(new_data)

Question 5

check for multicollinearity

Answer 5

variables = df[['col1','col2','col3']]
vif = pd.DataFrame()
vif["Features"] = variables.columns
vif["VIF"] = [variance_inflation_factor(variables.values, i) for i in range(variables.shape[1])]

Question 6

yhat equation from linear regression results

Answer 6

y_hat = x1*results.params[1]+results.params[0]

Question 7

logistic regression

Answer 7

x = sm.add_constant(x1)
reg_log = sm.Logit(y,x)
results_log = reg_log.fit()
results_log.summary()

Question 8

logistic regression predictions. Assume you have results in results_log

Answer 8

results_log.predict()

Question 9

logistics regression confusion matrix. Assume you have results in results_log

Answer 9

results_log.pred_table()