twidlr
twidlr copied to clipboard
drsimonj
data.frame-based API for model and predict functions
Just curious, but why would I use this package instead of the following... ``` df %>% lm(y ~ x, data = .) ``` I'm reading https://drsimonj.svbtle.com/a-tidy-model-pipeline-with-twidlr-and-broom and trying to figure...
Many twidlr methods involve setting and getting a formula as an attribute. It's a bit messy in base code. Eg... ``` attr(object, "formula") ``` Instead, create wrapper functions like `set_formula`...
Some models literally retain copies of data frames etc in order to make predictions. This can be convenient but has at least two downsides (described below). This issue proposes that,...
`predict` should always return a set of predicted values that match the data. twidlr can support various tidy functionality by always ensuring that the returned result is a vector or...
After having removed the .data.frame S3 methods for all model functions (see commit 0801907f13634c40bff403cc3577bf5031e0c93c), it occurs that S3 methods may not be necessary at all! Instead of having .default, just...
`predict.factanal` currently implements the "regression" method directly from `stats::factanal` (when `scores` = "regression"`). See https://github.com/drsimonj/twidlr/blob/master/R/twidlr-stats.R#L334-L340 `stats::factanal` also provides the means for computing factor scores using the Bartlett method, which can...
Predicting factor scores is a challenging issue. Existing approach makes use of code from stats::factanal that is partially copied and pasted. See: https://github.com/drsimonj/twidlr/blob/master/R/twidlr-stats.R#L334-L340 The stats authors must be cited appropriately.
The use of unexported/internal functions called via `:::` is not allowed by CRAN. For example, after running Rcmd check, a warning is produced: ``` Unexported objects imported by ':::' calls:...