| Re: PLOT() function THICK keyword in v8.2 just a guideline? [message #81309] |
Fri, 07 September 2012 07:23 |
David Fanning
Messages: 11724
Registered: August 2001
|
Senior Member |
|
|
Paul van Delst writes:
> O.k., I promise this is the last post (by me) on this since I have
> actual work work to do... I just can't help trying to figure out the
> level of busted-ness.
>
> Anyhoo, what I found out was:
>
> 1) the line thickness weirdness in the very first left-hand plot only
> manifests itself when the 4th left-hand plot (7th overall) is drawn.
>
> 2) if I comment out the use of the YRANGE keyword in the "pn" plot,
> everything works (except that sometimes the overplots extend beyond the
> plot boundaries since the plot range is determined by the first plot).
>
> So, somehow the yrange settings in the 4th left-hand plot is affecting
> the line thickness settings of the 1st left-hand plot, as well as all
> the remaining left-hand plots (where the yrange keyword is used).
>
> I'm going to burn some incense and hop in a counter-clockwise direction
> (on my left foot), all whilst mumbling oaths to the gods of gondwanaland.
>
> Maybe that'll fix it. :o)
You will recall my prediction that the function graphics system
is so ungodly and overly complicated that it will never work
properly. Fixing one thing breaks another.
I'm not saying this is what is happening here. But, as they say
about climate change, this pattern is consistent with what we would
expect IF that were happening! ;-)
Cheers,
David
P.S. I think the incantation requires you to also hold your nose
while you are hopping around.
--
David Fanning, Ph.D.
Fanning Software Consulting, Inc.
Coyote's Guide to IDL Programming: http://www.idlcoyote.com/
Sepore ma de ni thui. ("Perhaps thou speakest truth.")
|
|
|
|
| Re: PLOT() function THICK keyword in v8.2 just a guideline? [message #81310 is a reply to message #81309] |
Fri, 07 September 2012 07:18  |
Paul Van Delst[1]
Messages: 1157
Registered: April 2002
|
Senior Member |
|
|
O.k., I promise this is the last post (by me) on this since I have
actual work work to do... I just can't help trying to figure out the
level of busted-ness.
Anyhoo, what I found out was:
1) the line thickness weirdness in the very first left-hand plot only
manifests itself when the 4th left-hand plot (7th overall) is drawn.
2) if I comment out the use of the YRANGE keyword in the "pn" plot,
everything works (except that sometimes the overplots extend beyond the
plot boundaries since the plot range is determined by the first plot).
So, somehow the yrange settings in the 4th left-hand plot is affecting
the line thickness settings of the 1st left-hand plot, as well as all
the remaining left-hand plots (where the yrange keyword is used).
I'm going to burn some incense and hop in a counter-clockwise direction
(on my left foot), all whilst mumbling oaths to the gods of gondwanaland.
Maybe that'll fix it. :o)
cheers,
paulv
On 09/06/12 18:14, Paul van Delst wrote:
> Hello,
>
> I just got IDL v8.2 installed,
> IDL> print, !version
> { x86 linux unix linux 8.2 Apr 10 2012 32 64}
>
> I have a piece of code like that shown below (it's within a loop over
> index "i"). It simply reads some "new" data and some "old" data, then it
> plots the two of them together (the "pn" and "po" plots), and then it
> plots the difference (the "pd" plot):
>
> ; Create window to hold plots
> w = WINDOW( $
> WINDOW_TITLE = rts_file_list[i], $
> DIMENSIONS = [800,900], $
> BUFFER = buffer )
> ; ...Initialise layout index
> index = 0
>
>
> ; Loop over data items
> FOR j = 0, n_items-1 DO BEGIN
>
> ; Extract the current data item from the RTS objects
> new_data = new_rts.Get_Data(item_name[j], Profile=1)
> old_data = old_rts.Get_Data(item_name[j], Profile=1)
>
> ; Plot the two data sets
> index++
> yrange = [MIN(new_data)<MIN(old_data), $
> MAX(new_data)>MAX(old_data)]
> pn = PLOT( $
> new_data, $
> TITLE = item_name[j], $
> YRANGE = yrange, $
> LAYOUT = [ n_col, n_row, index ], $
> COLOR = 'red', $
> THICK = 2, $
> CURRENT = w )
> po = PLOT( $
> old_data, $
> COLOR = 'green', $
> OVERPLOT = w )
>
> ; Plot the difference
> index++
> pd = PLOT( $
> new_data-old_data, $
> TITLE = item_name[j]+" difference", $
> LAYOUT = [ n_col, n_row, index ], $
> CURRENT = w )
>
> ENDFOR
>
>
> Now, *most* of the time this code works fine. However, every now and
> again the line that is plotted by the "pn" plot, that is the one that
> has the keyword
> THICK=2
> plots the line with what looks like the equivalent of a thickness of
> about 10, and then some subsequent lines with a thickness of zero (i.e.
> not there, or hardly visible).
>
> It's as though the first super-thick line uses up all the thick,
> starving subsequent lines.
>
> Here's an example of the wrong result, with incorrect line thicknesses:
> http://tinyurl.com/bnchou4
>
> And here's an example of the expected result, with correct line
> thicknesses throughout:
> http://tinyurl.com/d394vsy
>
> I have the red line be twice as thick so I can see it behind the green
> one when the data are the same (too many points to use symbols).
>
> Is this a bug in IDL v8.2? Has anyone seen this behaviour before and
> (fingers crossed) discovered a solution?
>
> cheers,
>
> paulv
|
|
|
|
| Re: PLOT() function THICK keyword in v8.2 just a guideline? [message #81321 is a reply to message #81310] |
Thu, 06 September 2012 19:57 |
Russell Ryan
Messages: 122
Registered: May 2012
|
Senior Member |
|
|
On Thursday, September 6, 2012 7:11:38 PM UTC-4, David Fanning wrote:
> Paul van Delst writes:
>
>
>
>> But, seriously: sigh.
>
>
>
> Instead of a user's group, maybe we should form a support group.
>
>
>
> "Hi, my name is David and I'm an IDL graphics user."
>
>
>
> Cheers,
>
>
>
> David
>
>
>
>
>
> --
>
> David Fanning, Ph.D.
>
> Fanning Software Consulting, Inc.
>
> Coyote's Guide to IDL Programming: http://www.idlcoyote.com/
>
> Sepore ma de ni thui. ("Perhaps thou speakest truth.")
This basically sums up how I look and feel after a day with function graphics...
http://www.youtube.com/watch?v=ND31PWVW-TQ&feature=relat ed
russell
|
|
|
|
|
|
|
|
| Re: PLOT() function THICK keyword in v8.2 just a guideline? [message #81324 is a reply to message #81323] |
Thu, 06 September 2012 15:39 |
Paul Van Delst[1]
Messages: 1157
Registered: April 2002
|
Senior Member |
|
|
Oh boy. Further investigations reveal that, for plot with the incorrect
line thickness problem, when I give focus to a plot that has a
non-existant (or very thin) line and zoom in, the line thickness in
*all* the plots in the window suddenly appear as they should be.
When I click the "undo" button, the line thicknesses return to their
previous pear shaped wrongness. Unbelievable.
Did somebody at Excelis bury a whack-a-mole game inside the function
graphics byte code? (ha ha)
But, seriously: sigh.
cheers,
paulv
On 09/06/12 18:14, Paul van Delst wrote:
> Hello,
>
> I just got IDL v8.2 installed,
> IDL> print, !version
> { x86 linux unix linux 8.2 Apr 10 2012 32 64}
>
> I have a piece of code like that shown below (it's within a loop over
> index "i"). It simply reads some "new" data and some "old" data, then it
> plots the two of them together (the "pn" and "po" plots), and then it
> plots the difference (the "pd" plot):
>
> ; Create window to hold plots
> w = WINDOW( $
> WINDOW_TITLE = rts_file_list[i], $
> DIMENSIONS = [800,900], $
> BUFFER = buffer )
> ; ...Initialise layout index
> index = 0
>
>
> ; Loop over data items
> FOR j = 0, n_items-1 DO BEGIN
>
> ; Extract the current data item from the RTS objects
> new_data = new_rts.Get_Data(item_name[j], Profile=1)
> old_data = old_rts.Get_Data(item_name[j], Profile=1)
>
> ; Plot the two data sets
> index++
> yrange = [MIN(new_data)<MIN(old_data), $
> MAX(new_data)>MAX(old_data)]
> pn = PLOT( $
> new_data, $
> TITLE = item_name[j], $
> YRANGE = yrange, $
> LAYOUT = [ n_col, n_row, index ], $
> COLOR = 'red', $
> THICK = 2, $
> CURRENT = w )
> po = PLOT( $
> old_data, $
> COLOR = 'green', $
> OVERPLOT = w )
>
> ; Plot the difference
> index++
> pd = PLOT( $
> new_data-old_data, $
> TITLE = item_name[j]+" difference", $
> LAYOUT = [ n_col, n_row, index ], $
> CURRENT = w )
>
> ENDFOR
>
>
> Now, *most* of the time this code works fine. However, every now and
> again the line that is plotted by the "pn" plot, that is the one that
> has the keyword
> THICK=2
> plots the line with what looks like the equivalent of a thickness of
> about 10, and then some subsequent lines with a thickness of zero (i.e.
> not there, or hardly visible).
>
> It's as though the first super-thick line uses up all the thick,
> starving subsequent lines.
>
> Here's an example of the wrong result, with incorrect line thicknesses:
> http://tinyurl.com/bnchou4
>
> And here's an example of the expected result, with correct line
> thicknesses throughout:
> http://tinyurl.com/d394vsy
>
> I have the red line be twice as thick so I can see it behind the green
> one when the data are the same (too many points to use symbols).
>
> Is this a bug in IDL v8.2? Has anyone seen this behaviour before and
> (fingers crossed) discovered a solution?
>
> cheers,
>
> paulv
|
|
|
|
|
|
comp.lang.idl-pvwave archive
Members
Search
Help
Login
Home
