ggdag
ggdag copied to clipboard
r-causal
:arrow_lower_left: :arrow_lower_right: An R package for working with causal directed acyclic graphs (DAGs)
I think the coordinates API is unclear, as several people (myself included) have stumbled over it or missed it altogether. I think this needs a few changes: - [ ]...
In my own function '[plot.data.mappings](https://github.com/ben-oneill/utilities/blob/main/R/mappings.R)' I construct a DAG plot using ggplot2 with the ggdag functions for adding nodes, edges, etc. I have found that if you construct the plot...
As discussed in #41. Right now, a single `dagitty` object is associated with a single `tidy_dagitty` object, but it would be possible to manage multiple DAGs and map through them...
While #27 added basic tests for every R file, more extensive testing needs to happen, particularly testing function arguments, edge cases, and verifying output against `dagitty::fun()`
dagitty allows to specify a single control point when drawing edges, e.g. ``` dag { bb="-3,-0.5,2,1.2" D [outcome,pos="1.000,1.000"] E [exposure,pos="-2.000,1.000"] Z [pos="-0.500,0.500"] D Z [pos="1.000,-1.000"] E -> D E Z...
Thanks for the package. It took me some time to manage to use option `use_labels` in `ggdag`. Consider ``` dag % tidy_dagitty %>% ggdag(text = FALSE, use_labels = "label") ```...
ggraph 2.0.0 now uses `strength` instead of `curvature`. It may make sense to be consistent with this in ggdag
It would be nice to have a theme (or equivalent) to format the graph in the style of the DAGs in Causality (the book). Examples below. 
As discussed in #140, recent changes have broken the vignettes. They need to be updated to the newer syntactical changes.
With all the recent additions, I haven't been vigilant about test coverage. I should make sure it's high enough before 0.3.0