timedist should be able to fit multiple models

[nathaneastwood]

Currently you have to fit single models one at a time.

[nathaneastwood]

This is probably achievable with purrr

[nathaneastwood]

Actually, a much nicer way (and cleaner dependency way) would be to switch to data.table as the backend.

> dt <- data.table('x'=rnorm(100),'y'=rnorm(100),'g'=sample(c('a','b'),100,replace=T),key='g')
> GLMs <- dt[,list(glms=list(glm(formula=y~x))), by='g']
> GLMs[g == "a", glms][[1]]

Call:  glm(formula = y ~ x)

Coefficients:
(Intercept)            x  
    0.02669      0.16465  

Degrees of Freedom: 46 Total (i.e. Null);  45 Residual
Null Deviance:	    49.08 
Residual Deviance: 48.1 	AIC: 140.5