search

biocore / songbird

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

X axis of the CV error plots or tensorboard plots #159

Open manjumoorthy95 opened 2 years ago

manjumoorthy95 commented 2 years ago

From the readme of SongBird I can find that the X-axis of cverror graph is the number of iterations. In the command this is defined by "-epochs"_ argument. Though we had set the epochs to 10000, the X axis of the plots show the limits as 45K. Can you please explain this difference?

mortonjt commented 2 years ago

These are the iterations, there are often multiple iterations per epoch. This can be changed with the summary-interval argument.

On Wed, Mar 9, 2022 at 1:43 AM manjumoorthy95 @.***> wrote:

From the readme of SongBird I can find that the X-axis of cv_error graph is the number of iterations. In the command this is defined by "-epochs" argument. Though we had set the epochs to 10000, the X axis of the plots show the limits as 45K. Can you please explain this difference?

— Reply to this email directly, view it on GitHub https://github.com/biocore/songbird/issues/159, or unsubscribe https://github.com/notifications/unsubscribe-auth/AA75VXKDXGIGU7U3FOVRJALU7BCCZANCNFSM5QITOFUQ . Triage notifications on the go with GitHub Mobile for iOS https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675 or Android https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub.

You are receiving this because you are subscribed to this thread.Message ID: @.***>

manjumoorthy95 commented 2 years ago

Thanks for the clarification!

I had one more doubt. I wanted to create a model for checking differences between "Cancer and Control" while controlling for different other variables in the metadata. So is there a criterion that "Cancer and control" should be mentioned first in the formula?

Also please note that the "Cancer and Control" column lies the last in my metadata sheet, so will using this term in the first place of the formula cause any issues?

mortonjt commented 2 years ago

Hi @manjumoorthy95 the order of the terms in the formula won't matter.

athulkrishnacr commented 2 years ago

So, how do we control for the effect of certain variables ?

For example, if I want to know the differences between diseased states while controlling for 3 other variables, should the formula be like : " diseased_state+Var1+Var2+Var3" , where I am trying to control the effect of Var1 to Var3?