Add trend number to parameter calculation#45
Open
fnands wants to merge 1 commit intodiku-dk:masterfrom
Open
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Hi,
Unless I'm mistaken sigma is the residual standard deviation, so the denominator should be the d.o.f. i.e.
sample_size - number_paramswhich in this case will be2 * k + 1if there is no trend or2 * k + 2if there is a trend.It seems it was hardcoded to the case where there is a trend, so this should fix it.
I am unsure about the futhark implementation: it seems to into account the fact that there are fewer parameters when not calculating a trend, but frankly it's greek to me so I am unsure if it's actually being used.