[ncl-talk] gsLineThicknessF (bug?) in NCL 6.5

Tue Nov 20 11:40:36 MST 2018

Thank you so much Mary. I have followed your steps to report this issue on the github page, with your sample scripts and suggested fix.

The fix also solved my issue!

Nick Bassill, PhD

Modeler & Meteorologist, Center Of Excellence (Prior: NYS Mesonet)

LC SB-28, 1400 Washington Ave.,

SUNY, University at Albany, NY 12222

(518) 442-6375
NYSM Products: http://operations.nysmesonet.org/~nbassill/

________________________________
From: Mary Haley <haley at ucar.edu>
Sent: Tuesday, November 20, 2018 12:14 PM
To: Bassill, Nicholas
Cc: Ncl-talk
Subject: Re: [ncl-talk] gsLineThicknessF (bug?) in NCL 6.5

Hi Nick,

This is definitely a bug. I'll provide a work-around a little later in this email.

Meanwhile, as a favor, could you submit an issue using NCL's github issue tracker? We're moving from JIRA to github for issue tracking, and would like to get our users on this as well.

To report the issue:

[1] Go to the NCL github page:

https://github.com/NCAR/ncl

[2] Click on the "Issues" tab near the top of the page

[3] Click on the green "New Issue" button

[4] Fill in the title and description. The title should be descriptive, but as brief as possible. What you had for the subject of this email would work.

When filling out the description, you can use the attached sample script which illustrates the problem. Unfortunately, github doesn't let you select files with a ".ncl", but if you rename it to "xxx.ncl.txt" then you can attach it to the issue.

Or, you can copy the script, and then in the github description box, if you paste the code surrounded by these two lines:

```ncl
[....paste NCL code here...]
```

Then it will recognize this as an NCL script and will highlight bits of the code for you!  You can click on the "preview" tab to see what it looks like.

[5] Once the description is done, click on the "Submit new issue" button to submit it.

Back to the problem at hand. There are a couple of work-arounds.

Work-around #1

One is to fix the $NCARG_ROOT/lib/ncarg/nclscripts/csm/gsn_code.ncl script yourself.  Edit this line:

  gsres_tmp = get_res_eq(res2,"gs")           ; Get resource list for lines. Remove gsFill

and change "res2" to "res":

  gsres_tmp = get_res_eq(res,"gs")           ; Get resource list for lines. Remove gsFill

Work-around #2

If you don't want to edit and change  $NCARG_ROOT/lib/ncarg/nclscripts/csm/gsn_code.ncl directly, then I created a separate "gsn_table_fix.ncl" script for you, that has the new gsn_table with the one line fix.  You can include this fixed script in your current NCL script with:

load "./gsn_table_fix.ncl"

Let me know if this doesn't work for you.

--Mary

On Mon, Nov 19, 2018 at 12:14 PM Bassill, Nicholas <nbassill at albany.edu<mailto:nbassill at albany.edu>> wrote:

Hi All,

   I wanted to share a possible bug - although it's also possible I'm doing something wrong that I can't figure out. Long story short, I have a crontab'd, daily script that creates plots like this: http://operations.nysmesonet.org/~hudson2016/forecasts/2018103100SOMregressionforecast.png . That plot was created with NCL 6.3. It's a graph, with a table added to it below, with "res at gsLineThicknessF = 6" for the table line thickness. I relocated this script to a different computer that uses NCL 6.5. That plot looks like: http://operations.nysmesonet.org/~hudson2016/forecasts/2018110100SOMregressionforecast.png - notice the table line thicknesses are much thinner.

    I assumed I changed something in the move (and forgot), but recently I started making a new plot, also with a table attached to the bottom, using NCL 6.5, and sure enough, I can't change the line thickness.

Even though I know most of these resources aren't relevant, these are the resources I've rotated through trying:

  tableres at gsLineThicknessF = 10.0
  tableres at gsFillLineThicknessF  = 25.0
  tableres at gsEdgesOn = True
  tableres at gsEdgeThicknessF = 20.0

 .... all with no luck. I'm wondering if anyone else has experienced this issue, or has any suggestions?

My script is very lengthy, and the underlying data complex, so I'm not sharing it here, but the table portion looks like:

tableres = True
  ncr = (/11,4/)     ; 4 rows, 3 columns
  tableres at txFontHeightF = .012 ; fontvalues       ; Font height
  tableres at gsLineThicknessF = 10.0
  tableres at gsFillLineThicknessF  = 25.0
  tableres at gsEdgesOn = True
  tableres at gsEdgeThicknessF = 20.0

  tableres at txJust = "CenterCenter"
ncr = (/4,11/)
tableres at txFont = "helvetica-bold"
  text := round(transpose(forecastoutages),3)
  text := where(text.lt.0,0,text)+""

  y := (/0.02,0.12/)         ; Start and end X
  x := (/0.2,0.8/)         ; Start and end Y

  tableres at txFontHeightF = .011 ; fontvalues       ; Font height
  tableres at txJust = "CenterCenter"

  colors := 10+round(transpose(forecastoutages),3)
  colors := where(colors.gt.220,220,colors)
  colors := where(colors.lt.10,10,colors)

  tableres at gsFillColor := colors

  gsn_table(wks,ncr,x,y,text,tableres)

ncr = (/4,1/)
x = (/.04,.2/)
res at gsFillColor := "white"
gsn_table(wks,ncr,x,y,(/"10% Worst Case:","Best Forecast~N~:","90% Worst Case:","99% Worst Case:"/),tableres)

ncr = (/1,11/)
x = (/0.2,0.8/)
y = (/0.12,0.145/)

  gsn_table(wks,ncr,x,y,dayarray,tableres)

Nick Bassill, PhD

Modeler & Meteorologist, Center Of Excellence (Prior: NYS Mesonet)

LC SB-28, 1400 Washington Ave.,

SUNY, University at Albany, NY 12222

(518) 442-6375
NYSM Products: http://operations.nysmesonet.org/~nbassill/
_______________________________________________
ncl-talk mailing list
ncl-talk at ucar.edu<mailto:ncl-talk at ucar.edu>
List instructions, subscriber options, unsubscribe:
http://mailman.ucar.edu/mailman/listinfo/ncl-talk
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.ucar.edu/pipermail/ncl-talk/attachments/20181120/f60af144/attachment.html>