Save/capture message "Notice: Data set contains sampled data"

[efh0888]

Is it possible to somehow capture this message, i.e. store it as a character string for later use? I've got a script that automates our data extracts from Google Analytics using the rga package, and a table that keeps logs for each query, so it'd be nice to include a column there indicating whether or not sampling was used.

[mattpolicastro]

The Reporting API response does include fields regarding sampling (containsSampledData, sampleSize, and sampleSpace) so it's not unfeasible. How would you see this data being returned, though? As another column on the results table?

[MarkEdmondson1234]

I put this data as an attribute of the data frame results, so you can reach it via say attr("sampled", garesults)

[efh0888]

I think an attribute makes sense

[mattpolicastro]

Had no idea attributes were a thing. Seems great. Prefix the attributes as something like ga:containsSampledData to prevent collisions?

[mattpolicastro]

Took a first stab at this with a simple conditional block in core.R. Thoughts?

[efh0888]

Looks good to me, except the additional warning is redundant. Might also want to add something like

else {
  attr(ga.data.df, "ga:containsSampledData") <- FALSE
}

[mattpolicastro]

Yep, the warning was an oversight on my part. Latest commit.

I will say I'm having some funny feelings about prefixing the attributes with "ga:"—I like it for clarity and (hopefully) avoiding collisions upstream, but it's fairly indelicate.

[MarkEdmondson1234]

I would avoid the ':' as it may do funky things to R names.

As a suggested enhancement, I got this from RGoogleAnalytics and it may be good here - it replicates the "This data is drawn from 2% of visits" messages you get in the GA interface:

samplePercent <-  100
  if(!is.null(x$containsSampledData)){
    if(x$containsSampledData) {
      samplePercent <- round(100 * (as.numeric(x$sampleSize) / as.numeric(x$sampleSpace)), 2)
      message("Data is sampled, based on ", samplePercent, "% of visits. Use WALK to mitigate it." )
    }
  }

[mattpolicastro]

I like it. I'll add something in.

[EricGoldsmith]

FWIW, I did something similar in my local fork (and, apologies, never submitted a PR): https://github.com/tedconf/rga/commit/74dd9a44c5d31d7c99657d3f7a4f15136ad7741e

[mattpolicastro]

@EricGoldsmith yeah, that's about what I had in mind. Mind if I pull it into my fork so I can PR it?

[EricGoldsmith]

@mattpolicastro, be my guest.

[mattpolicastro]

Super. I'll restrict myself to that particular commit as your other changes won't merge as cleanly, but I'll start in on it.

[mattpolicastro]

Ok, I think I've incorporated everything into one branch. Any more thoughts before I submit a PR?

[MarkEdmondson1234]

:+1:

[EricGoldsmith]

@mattpolicastro, can you confirm that this is now fixed, and then close this issue?

[mattpolicastro]

Sorry for not closing the loop on this—it appears to be working correctly after force-updating the package install. However, it looks like @efh0888 may have to close it as I'm not an admin.