You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Occurs to me a simple version of this is dots/ellipses or squares that have different vertical sizes (easier layout problem than circles that have different radii, and more correct anyway). Something like a geom_weighted_dots(). Would need to do a bunch of stuff:
finally implement use of arbitrary layout algorithms in the binwidth finding algorithm (e.g. a shortcut version of each layout alg that just returns max height)
possibly allow limited support for shapes (at least ellipses, squares, maybe rounded rectangle or a pill shape, maybe even reimplement all pch shapes manually but allow stretching, or at least use the affine transforms to do it).
implement a geom_weighted_dots() and stat_weighted_dots() (for rvars at least).
possibly give a warning (maybe with option to turn it off --- or just tell people to unset the weights aesthetic) if weights are used with geom_dots(), suggesting folks used geom_weighted_dots()?
also consider if this should be a more visual aesthetic (like "dotheight" or "dotextent" or ???) that weights from a dist could be mapped onto?
The text was updated successfully, but these errors were encountered:
Occurs to me a simple version of this is dots/ellipses or squares that have different vertical sizes (easier layout problem than circles that have different radii, and more correct anyway). Something like a
geom_weighted_dots()
. Would need to do a bunch of stuff:geom_weighted_dots()
andstat_weighted_dots()
(forrvar
s at least).geom_dots()
, suggesting folks usedgeom_weighted_dots()
?The text was updated successfully, but these errors were encountered: