Linear Models

Question 1

What is the equation used to demonstrate a linear relationship

Answer 1

response ~ explanatory

Question 2

What are the three parts of ols

Answer 2

1. Formula
2. Fit
3. Summary

Question 3

Why do we test data for normal distribution?

Answer 3

Certain statistical tests assume that data is normally distributed, such as regression analysis and t-tests. If the data is not normally distributed, than non-parametric tests (that dont rely on the normal curve of distribution) can be used like the Mann-whitney U test. OR MORE COMMONLY we will log transform the data.

Question 4

How can you tell if data is normally distributed?

Answer 4

Shapiro-Wilk Test: Tests the null hypothesis that the data is normally distributed. A p-value > 0.05 indicates the data is likely normal.

stat, p_value = shapiro(dataframe[new_col_name])
print(f’Shapiro-Wilk test for {new_col_name}:’)
print(f’Statistic={stat}, p-value={p_value}\n’)

Question 5

How do you log-transform data?

Answer 5

def log_transform_and_test(dataframe, columns):
# Apply log transformation to specified columns and add new columns to the dataframe
for col in columns:
new_col_name = f’log{col}’
dataframe[new_col_name] = np.log(dataframe[col])