Step 4 By default, statsmodels OLS will estimate one parameter or coefficient for each independent variable. In our case, we have one independent variable (x), and we want to estimate the linear coefficient that relates x and y, which we are calling . But we also want OLS to estimate the y-intercept 0 . In order to do so, we must augment the dataset by adding an additional column to the data in x so that OLS can separately estimate a y-intercept from that. The mechanism that statsmodels uses to trigger this estimation is to add a constant column of ones to the independent variable, which OLS then interprets as a constant offset to the line it is fitting. You might find it useful to consult the statsmodels documentation on OLS. In the code cell below: Use the sm.add_constant function to add a constant to x, and assign the result back to x Print the variable x. You should note that this now contains two columns, one labeled const (the constant we just added) and one labeled LogGDP