3D plot showing the regression model solution represented by a plane. The plane is the representation of the points (z^) that best fit the observations (z). It is the solution of a theoretical exercise that seeks to understand sales by setting two explanatory variables: price and investing in advertisement. The model (i.e., the plane) can explain or predict sales as a function price and sales where each beta (b1, b2, and b3) define the plane. The plane is oriented such that the errors are minimum.
Green dots (z) = actual observations (sales)
Black dots (z^) = estimated values knowing x and y (price and advertisement), which are defined by z^=b1 + b2(price)+b3(advertisement)
Redlines = The error (residual or e) between the observed sales (z) and the estimated sales (z^)
Yellow triangles = Show the change in z holding the other variables constant.
The points were developed from a .csv file and uploaded in Blender using python. Then formatted in Scketchfab Provide some feedback so it can improve
3 comments