Tim Taylor

The branch https://github.com/reconhub/trendbreaker/tree/trending now has compatibility with . `asmodee` altered to a generic with methods for `data.frame` and `incidence2`. Plot also modified to work with grouped incidence objects.

actually, probably better to convert to character and let glm do the failing...

``` r library(trendbreaker) glm_model #> Coefficients: #> (Intercept) cyl drat wt qsec am #> 6.256607 0.085193 -0.009274 0.176041 -0.136688 0.036604 #> #> Degrees of Freedom: 31 Total (i.e. Null); 26...

will be closed by [trending](https://github.com/reconhub/trending)

Hi @andrew-farrey, I'm guessing you've recently updated or reinstalled {trending} or {trendeval} and possible also {incidence2}? New versions of these were recently released to CRAN and we will need to...

Currently looking at this in the [noci](/reconhub/trendbreaker/tree/noci) branch.

If we avoid concatenation/rep we can eek out a little more performance. Ignoring the method side of things for the moment but to illustrate the gain ... ``` r duplicated_numeric_version_2

will be closed by [trending](https://github.com/reconhub/trending)

Cheers - Sorry for being the (pesky) user wanting to ruin the beautiful code! I appreciate the thought you're putting in to this! Another benefit from the user perspective is...

Cheers @avallecam. I'm going to ponder this for a while as I do understand this may be desirable. The current design is due to me wanting to encourage people to...