search

ISA-tools / stato

This is the development repository for the STATistics Ontology (STATO). For more information and demonstration on the ontology content, please visit its website:
http://stato-ontology.org/
31 stars 8 forks source link

"contrast weight estimate" re-labelled "contrast estimate"? #34

Closed cmaumet closed 9 years ago

cmaumet commented 9 years ago

Hi,

Following #23, the term "contrast weight estimate" was created in stato. I wonder if this term would be better labelled "contrast estimate" as it is not the weights themselves that are estimated but the contrast defined by them. @nicholst: would you agree with that?

How would you feel about this?

(We could then re-use this term in relation to nidm:ContrastMap).

nicholst commented 9 years ago

Yes, I concur with this. As previously discussed, there is much ambiguity with contrasts in general, but with "contrast estimate" there is no ambiguity, and it is clear that we're taking about c\hat{\beta}, ie the weighted sum of estimated parameters as specified by the "contrast weights". (And as the contrast weights c are always known and never estimated, there is no ambiguity there.

On Feb 27, 2015, at 10:34 AM, Camille Maumet notifications@github.com wrote:

Hi,

Following #23, the term "contrast weight estimate" was created in stato. I wonder if this term would be better labelled "contrast estimate" as it is not the weights themselves that are estimated but the contrast defined by them. @nicholst: would you agree with that?

How would you feel about this?

(We could then re-use this term in relation to nidm:ContrastMap).

— Reply to this email directly or view it on GitHub.

agbeltran commented 9 years ago

Hi @cmaumet & @nicholst,

I see what you mean about re-labelling, but am I right that the PR #37 creates a new term instead of re-labelling the existing 'contrast weight estimate' (STATO_0000324)?

Thanks

cmaumet commented 9 years ago

Hi @agbeltran! Yes, you are right, #37 creates a new term for "contrast estimate" rather than renaming "contrast weight estimate" as suggested here. I was not sure what you wanted to do with "contrast weight estimate" so I just thought I would rather create a separate term... (Also, I don't think a "contrast estimate" would be a child of "model parameter estimate" so this ended up being more than just re-labelling...).

If that's fine with you, I am happy to close this issue and discuss at #37 instead?

cmaumet commented 9 years ago

I am closing this issue as this has been discussed and updated in the merged pull request #37.