When constructing a train/test split in SAS, a data scientist sets the random seed before calling PROC SURVEYSELECT. What is the main purpose of setting the seed?
Select an answer to reveal the explanation.
Short Explanation and Infographic
Here's the deal — b is correct because a fixed random seed initializes the pseudo-random number generator at the same starting point, producing the same split on every run and enabling reproducibility. A describes stratified sampling, which is a separate option in PROC SURVEYSELECT.
Full explanation below image
Full Explanation
B is correct because a fixed random seed initializes the pseudo-random number generator at the same starting point, producing the same split on every run and enabling reproducibility. A describes stratified sampling, which is a separate option in PROC SURVEYSELECT. C describes oversampling or class weighting, not seeding. D describes CAS parallelism, which is unrelated to the seed value.