How can use linear model for mutiple columns for forecasting?

Question

I am trying to create a user defined function to forecast multiple time series using linear regression? I got stuck don't know what happened.. Here is my code: import statsmodels.formula.api as smf lst=[3,4,5,6] def linear(lst): for i in lst: #model=smf.ols('Footfalls~i',data=Train).fit() exp=smf.ols('Footfalls~Walmart[i]',data=Train).fit() return (pd.Series(linear_model.predict(pd.DataFrame(Test['t']))))

Data Set:

Month Footfalls t t2 log t

Jan-19 1707 1 1 7.4 Feb-19 1235 2 4 7.3 Mar-10 5677 3 9 7.5

I have a similar data where i want to built a regression model for t,t2,logt.