IDEMSInternational / R-Instat

A statistics software package powered by R
http://r-instat.org/
GNU General Public License v3.0
38 stars 102 forks source link

Corrections on plotting sub-dialog and on line plots #8992

Closed rdstern closed 2 weeks ago

rdstern commented 1 month ago

@MeSophie the first three of these should be simple to fix? I hope you can be doing these tasks while you are still completing your current dialogs, where you sometimes seem a bit stuck. (I'm keen for you to return to your earlier higher productivity level!?)

a) In the plotting sub-dialog on the x-axis and y-axis I would often want to set the x-axis lower limit to go down to zero. I don't want to specify the upper limit.

image

Currently, if I do this, then the change is ignored. If I give both limits then it is fine. The code is then limits = c(0,10) say. Please could you also permit the limits to be given if just one is specified. It seems to work with c(0, NA), or c(NA, 10), so that's what the code should do.

b) In the line plot there is (I think reasonably recently) an Orientation parameter. Could you please add that into geom_line. I think the default is NA or "y" or "x". Please check and try. (It is in the tidyverse example script.)

c) There is also an Arrow parameter. That is given nicely in geom_path. Please improve it there a bit and also add it to geom_line, in the same position. The obvious improvements are to give some options. Currently there is a drop-down, but with the single option arrow( ). Keep that as the default and add arrow(angle=15), arrow(ends="both") and arrow(type="closed"), and arrow(angle=15, type = "closed") as further options. (There's many more that are possible, but this is enough for now to indicate the range of options.

d) (This may take longer, unless it is obvious - it may be!) I used the line plot quite a lot and it seemed fine. Then I used the smoothing option. After that I went back to the line options and found the smooth stuff remained included. I hope you can check and find that this is an isolated event of poor coding and it not indicative of lots of changes being needed?