biocore / songbird

Vanilla regression methods for microbiome differential abundance analysis
BSD 3-Clause "New" or "Revised" License
58 stars 25 forks source link

More Patsy formula examples in README? / More documentation on design matrices, ANOVA, etc. #81

Open fedarko opened 5 years ago

fedarko commented 5 years ago

Something I've been thinking about is that, to a pretty large degree, "using Songbird" boils down to choosing a formula.

I know that the current README already links to the Patsy docs, and I know that explaining statistical formulas to the user in the README is a tall order, but it might be a good idea to at least add a few example formulas for common use cases (e.g. using C() and Treatment coding to explicitly set a reference).

...At the very least we could add a link to this blog post :)

mortonjt commented 5 years ago

It'll take some time, it's basically a language of its own.

There is also this blog post on how to do ordinal variables

https://mortonjt.github.io/probable-bug-bytes/probable-bug-bytes/ordinal-variables/

On Sat, Sep 21, 2019, 2:45 AM Marcus Fedarko notifications@github.com wrote:

Something I've been thinking about is that, to a pretty large degree, "using Songbird" boils down to choosing a formula.

I know that the current README already links to the Patsy docs, and I know that explaining statistical formulas to the user in the README is a tall order, but it might be a good idea to at least add a few example formulas for common use cases (e.g. using C() and Treatment coding to explicitly set a reference).

...At the very least we could add a link to this blog post http://mortonjt.blogspot.com/2018/05/encoding-design-matrices-in-patsy.html :)

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/biocore/songbird/issues/81?email_source=notifications&email_token=AA75VXMLDWU7VXAVXUMXZY3QKW7HRA5CNFSM4IY5GDCKYY3PNVWWK3TUL52HS4DFUVEXG43VMWVGG33NNVSW45C7NFSM4HMZ3G4A, or mute the thread https://github.com/notifications/unsubscribe-auth/AA75VXIO4N7JP4GKLIX7W5LQKW7HRANCNFSM4IY5GDCA .

fedarko commented 5 years ago

I didn't know you had a post on ordinal variables—that's super helpful, thanks!

I understand that going super in-depth probably isn't a great use of time, since patsy is really comprehensive. I did add a small "Specifying a formula" section to the PR's README, but it's relatively small in scope -- it includes:

This should at least be a good start for many use cases?

mortonjt commented 5 years ago

yes! this is awesome!

On Sat, Sep 21, 2019 at 9:18 PM Marcus Fedarko notifications@github.com wrote:

I didn't know you had a post on ordinal variables—that's super helpful, thanks!

I understand that going super in-depth probably isn't a great use of time, since patsy is really comprehensive. I did add a small "Specifying a formula" section https://github.com/fedarko/songbird/#3-specifying-a-formula to the PR's README, but it's relatively small in scope -- it includes:

  • the prior FAQs on formulas
  • links to patsy's docs
  • links to your blog posts on Treatment and Backwards Difference coding
  • small examples of Treatment and Backwards Difference coding, based on your blog posts

This should at least be a good start for many use cases?

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/biocore/songbird/issues/81?email_source=notifications&email_token=AA75VXPUVRVC6ZKFRN6Z7JDQK3BXJA5CNFSM4IY5GDCKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOD7I4OLQ#issuecomment-533841710, or mute the thread https://github.com/notifications/unsubscribe-auth/AA75VXOPB5EZM4WXQCIXXXTQK3BXJANCNFSM4IY5GDCA .