Prepare Data for Random Forest | Dataford Interview Questions

Given this dataset with missing and anomalous values, write a script to prepare it for a random forest classifier.

Implement prepare_dataset(rows, target_column), where rows is a list of dictionaries containing numeric, categorical, or None values. Return a dictionary with one-hot-encoded feature rows and labels. Treat numeric values outside the 1.5 IQR range as missing, replace missing numeric values with the median, replace missing categorical values with the mode, and omit rows whose target is missing. Preserve the target values unchanged.

Constraints

1 <= rows.length <= 10^4
Each row contains the same feature keys and one target key
Feature values are None, numeric values, strings, or other hashable categorical values
At least one nonmissing value exists for each feature column
Rows with a missing target are excluded from the output

Function Signature

def prepare_dataset(rows, target_column):

Your question is Prepare Data for Random Forest. Start with the requirements in the Question tab.

Run and submit as often as you like. When you're ready, talk me through your approach or go straight to the code.

Problem

Given this dataset with missing and anomalous values, write a script to prepare it for a random forest classifier.

Constraints

1 <= rows.length <= 10^4
Each row contains the same feature keys and one target key
Feature values are None, numeric values, strings, or other hashable categorical values
At least one nonmissing value exists for each feature column
Rows with a missing target are excluded from the output

Function Signature

def prepare_dataset(rows, target_column):

Your question is Prepare Data for Random Forest. Start with the requirements in the Question tab.

Run and submit as often as you like. When you're ready, talk me through your approach or go straight to the code.