Pretend Everyone Had It — easy Machine Learning in Code problem | Incognition

Pretend Everyone Had It

10 pts · 10 coins

Machine Learning in CodeEasy

Pretend Everyone Had It

Partial dependence is a thought experiment you can run yourself: give every row the same value in one column, ask the model for its predictions, and average them. Repeat for each value you want to try.

A letting agent predicts monthly rent from [floor_area_m2, building_age_years].

Task: write average_prediction_at(rows, rents, column, values).

Fit LinearRegression() on rows and rents.
For each number in values, in order: copy the rows, set column column of every copied row to that number, predict all the copied rows, and take the mean prediction.
Return the means as plain floats, each rounded to 2 decimal places.

Examples

rows = [[102.0, 37.0], [47.0, 38.0], [74.0, 8.0], [94.0, 13.0], [119.0, 13.0], [96.0, 14.0], [116.0, 21.0], [92.0, 27.0], [61.0, 29.0], [89.0, 26.0]]

rents = [308.5, 152.6, 263.4, 328.0, 412.1, 338.9, 387.5, 304.4, 211.0, 300.9]

column = 0

values = [50, 80, 110]

returns [180.76, 273.04, 365.33]

rows = [[102.0, 37.0], [47.0, 38.0], [74.0, 8.0], [94.0, 13.0], [119.0, 13.0], [96.0, 14.0], [116.0, 21.0], [92.0, 27.0], [61.0, 29.0], [89.0, 26.0]]

rents = [308.5, 152.6, 263.4, 328.0, 412.1, 338.9, 387.5, 304.4, 211.0, 300.9]

column = 1

values = [5, 50]

returns [325.21, 262.62]

Hints

solution.py

1import numpy as np
2from sklearn.linear_model import LinearRegression
3
4
5def average_prediction_at(rows, rents, column, values):
6    pass
7