Generating data
Now that we've gone through the process of searching for the right model, let's talk about what the model's true parameters were and how they line up with the parameters our regression generated.
This is the code that was used to generate the data:
import numpy as np variable_a = np.random.uniform(-100, 100, 30) variable_b = np.random.uniform(-5, 5, 30) variable_c = np.random.uniform(0, 37, 30) variable_d = np.random.uniform(121, 213, 30) variable_e = np.random.uniform(-1000, 100, 30) variable_f = np.random.uniform(-100, 100, 30) variable_g = np.random.uniform(-25, 75, 30) variable_h = np.random.uniform(1, 27, 30) independent_variables = zip(variable_a, variable_b, variable_c, variable_d, variable_e, variable_f, variable_g, variable_h) dependent_variables = [3*x[0] - 2*x[1] - .25*x[4] + 5.75*x[1]*x[2] + np.random.normal(0, 50) for x in independent_variables] full_dataset = [x[0] + (x[1],) for x in zip(independent_variables, dependent_variables)] import csv with open('generated_data...