Lab 9: Regression (optional)#

Caution

This labb will provide you with an opportunity to practice regression modeling in Python. We will be working with Guerry moral statistics data, that we already came across when studying LISA. Please start by opening this Google Colab notebook and copying it to your Google Drive.

Important

Please submit the .ipynb notebook by modifying the code from the Google Colab notebook above.

Warning

Make sure to submit the lab by November 28 / 29, 2022, depending on the scheduled time of your section.

Lab Instructions#

  1. Check the normality of your outcome variable MEDV (hint: use a histogram). Create a correlation matrix for all of the variables.

  2. Fit the linear regression to the data using the following predictors: CRIM, ZN, INDUS, CHAS, NOX, RM, AGE, DIS, RAD, TAX, PTRATIO, B, LSTAT to model MEDV. We will refer to this model as m1.

  3. Report and interpret your R-squared and Adjusted R-squared. Comment on the coefficients and denote which ones are significant and insignificant (given their p-values)

  4. Now fit only predictors that only had significant p-values (call this model m2).

  5. Report and interpret your R-squared and Adjusted R-squared. Comment on the coefficients and denote which ones are significant and insignificant (given their p-values).

  6. Compare m1 to m2. Which factors appear to be most significant for MEDV?

  7. Calculate and report Global Moran’s I for m2 residuals.

  8. Plot m2 residuals versus the lag of residuals.

  9. Calculate Local Moran’s I and plot it on the map. Is there any spatial structure to m2 residuals?

  10. Fit a fixed effect model m3. Report which neighborhood has the highest and lowest coefficient. Interpret. Report and interpret your R-squared and Adjusted R-squared. Comment on the coefficients and denote which ones are significant and insignificant (given their p-values).

  11. Run the spatial error model m4. Report and interpret pseudo R-squared (how does it compare to m1, m2, m3?) Comment on the coefficients and denote which ones are significant and insignificant (given their p-values).

  12. Submit to GauchoSpace as ipynb/pdf.