Pedro J. Aphalo

@slowkow Commit 443105a3e21eb2693c5d6235dca8195f4cf6e71e in #196 fixes 1. and 3. The causes were multiple, I briefly explain the new code in a comment. 2. The problem with repulsion starting towards the...

What looks like a similar problem affecting justification can be seen in some of the examples of the new position nudge functions.

@slowkow This problem seems to be related to an interaction between justification and box padding. It still exists in the code in pull request #196 as of commit 76d8ff4ad889ddba8950321023cbee198601809c ```...

At least for this simple example adding radial nudging away from the origin gives the desired result. NOTE: This is using ggrepel under development and ggpp (>= 0.4.2) from CRAN....

This can be solved using nudging. Function `positiion_nudge_to()` from package ggpp solves this problem. It takes a vector of nudgings, so by manually or programatically computing these distances you can...

There seems to be an additional problem when both `hjust` and `vjust` are passed to `geom_text_repel()` adding nudging even when nudge is set to zero. In these examples `"right"` and...

@slowkow It took me quite a lot of trial an error to get the justification working in ggplot2. I started poking in `geom_text()` when I started seeing odd behavior from...

@slowkow #196 updates `compute_just()` and deals only with the handling of specific values of angle together with `"inward"` and `"outward"` justification. So, even though this change can affect the rendering...

@slowkow #196 seems to correct the "inward"/"outward" swap problem when `angle != 0`. One can also see that when angle takes certain values the segment is not clipped to the...

@glorious1 Using values outside the range 0..1 for justification is problematic. You may want to use nudging and justification together. You could disable repulsion if you have manually set the...