helpVeusz - A Scientific Plotting Package - Support: sr #2725, signed descriptors as column titles

 
 
Show feedback again

sr #2725: signed descriptors as column titles

Submitted by:  None
Submitted on:  Mon 28 Mar 2011 09:42:19 PM UTC  
 
Category: NonePriority: 5 - Normal
Severity: 3 - NormalStatus: None
Privacy: PublicAssigned to: None
Originator Email: -unavailable-Open/Closed: Open
Operating System: GNU/Linux

Add a New Comment (Rich MarkupRich Markup):
   

You are not logged in

Please log in, so followups can be emailed to you.

 

(Jump to the original submission Jump to the original submission)

Wed 30 Mar 2011 09:04:40 PM UTC, comment #8:

Getting rid of those spaces did the trick, thanks. I now realize that what I've been trying to do is a non-starter: I had assumed Veusz automatically accepts/updates the column name to whatever it finds at the head of the particular column one had (upon first setting up the .vsz file) elected as (e.g.) "Y data," and that one would not have to go through the step of manually making the associations each time the descriptor line changes.

John Wilson <jdwilson>
Wed 30 Mar 2011 08:30:00 PM UTC, comment #7:

Your recent example file seems to work for me, but I'm not aware of any changes I've made to the program recently which would affect this. I could plot uonsigu against the other datasets, like A-0.400.

If you type the name by hand you need to have the right number of spaces after the dataset name (those within ``). It might be best to not include spaces after the text within `` when you output the data file in fortran.

I'll attach my example vsz file showing it is working.

(file #12747)

Jeremy Sanders <jeremysanders>
Project Administrator
Wed 30 Mar 2011 08:04:11 PM UTC, comment #6:

Thanks Jeremy. I've tried the `A-0.500` notion as descriptor, and the file is attached. Veusz reads in the file quite as expected, knows each column has 241 rows (not counting the descriptor row), shows the descriptors correctly in the dialog when setting up an xyplot, but doesn't plot anything. Selecting uonsigu as descriptor for the X data immediately causes the X axis range to adjust, correctly reflecting that column of values whose header is uonsigu; however when it comes to setting up the Y data, no matter which of the columns that are headed by `A...` one chooses, there is no response -- the Y-axis range sites at 0-1. If your version plots this file correctly then I guess I do need to update.

About the interface, I'd happily write a help page but at this stage I'm too confused myself. I'm not clear in my own mind what the distinction is between the mailing list and the "github bug tracker" nor which it is I am using. No doubt in time it will become clear, though I wouldn't like to understate my ability to rapidly forget things I'm not constantly using! John.

(file #12745, file #12746)

John Wilson <jdwilson>
Wed 30 Mar 2011 07:29:46 PM UTC, comment #5:

I think you need the backticks in the file to import the data, which seems to work. There's no way to distinguish the error bars from the dataset names otherwise. It's possible the descriptor parser could be made more intelligent however, but people might rely on the ability to do "A+-"

Veusz lets you enter numbers manually in the xy plotter in the dataset box, so that's why it's getting confused by the dataset names. It should allow you to also enter names which look like numbers using backticks but it doesn't work currently. I should investigate that bug.

You could do a workaround by putting a letter before the - signs, e.g.

descriptor `A-1` `A-2` `A+2` `A+1`
1 2 3 4
4 5 6 7

This works okay for me.

I'm sorry about this pretty terrible interface - there's not much to suggest other than use the mailing list or the github bug tracker (linked from the veusz webpage).

Jeremy Sanders <jeremysanders>
Project Administrator
Wed 30 Mar 2011 06:54:30 PM UTC, comment #4:

Oh good -- that worked. The attached files are bigger versions than those I clipped into the message earlier. Each has a different descriptor line. Only rampCKLS_generic.dat plots properly (in that case I did not attempt to provide numeric descriptors). Thanks again, John.

John Wilson <jdwilson>
Wed 30 Mar 2011 06:50:00 PM UTC, comment #3:

Realize I could have attached files. Without starting another "Post a Comment" I used the Browse boxes under the panel "Attached Files" to insert several file names. A message in red continues to inform me "No files currently attached." Not obvious to me what I'm supposed to do. I'll try Submit Changes

(file #12742, file #12743, file #12744)

John Wilson <jdwilson>
Wed 30 Mar 2011 06:35:50 PM UTC, comment #2:

Hi.. I'll attach two examples. In the first I have converted numbers to strings (in fortran), prepended and appended back quotes, and written to the data file. In this case the minus signs are not interpreted as signifying an error column, and the descriptors correctly show up as available column headers when one is looking an the x-y plot widget (Properties x-data, y-data) but the columns do not plot as expected. When I looked at this yesterday with a larger file, it appeared the descriptor was being interpreted as being (itself) the sought for data, a scalar in lieu of a vector -- the scale on the Y-axis would adjust and a single point was plotted at the value of Y equal to the (numeric value corresponding to) the descriptor.

In the second file below the descriptors were written out from fortran as numbers. The negative descriptors do not show up as column headers when one is assigning the properties.

Data file 1 (there were actually more columns):
descriptor uonsigu `-0.500 ` `-0.250 `
-6.0000 0.59918E-08 0.59924E-08 0.46909E-06
-4.0000 0.13195E-03 0.13196E-03 0.62428E-03
-2.0000 0.53224E-01 0.53227E-01 0.61365E-01
0.0000 0.39329 0.39328 0.39410
2.0000 0.53226E-01 0.53227E-01 0.43911E-01
4.0000 0.13195E-03 0.13196E-03 0.47163E-05

Data file 2:
descriptor uonsigu -0.5000 -0.2500 0.000
-6.0000 0.59918E-08 0.59924E-08 0.46909E-06
-4.0000 0.13195E-03 0.13196E-03 0.62428E-03
-2.0000 0.53224E-01 0.53227E-01 0.61365E-01
0.0000 0.39329 0.39328 0.39410
2.0000 0.53226E-01 0.53227E-01 0.43911E-01
4.0000 0.13195E-03 0.13196E-03 0.47163E-05
6.0000 0.59920E-08 0.59923E-08 0.91002E-12
6.0000 0.59920E-08 0.59923E-08 0.91002E-12

What is being plotted is a probability density function for (normalized) velocity, for discrete values given in column 1 (uonsigu). Columns 2 ... give the pdf at a series of points in space. The coordinates of those points are used as the descriptor, at least that is the objective. Thanks, John.

PS. I'll await comment on this before trying out the test release.

John Wilson <jdwilson>
Wed 30 Mar 2011 09:15:15 AM UTC, comment #1:

Hi - Would it be possible to attach an example file? Is this a CSV file with something like

1 -3 -2 -1 1 2 3
2 -4 -2 -1 4 6 7
3 ...
4
5

?

If so, this seems to work ok for me in Veusz, where the columns get called col1, col2, etc, and contain the negative values. It's possible this is broken in 1.10 and fixed in the development version, so you could try the test release in http://download.gna.org/veusz/testing/

Jeremy Sanders <jeremysanders>
Project Administrator
Mon 28 Mar 2011 09:42:19 PM UTC, original submission:

Hello Jeremy,

Suppose one had vectors of length (say) i=1...10, and a column-organized file whose first column is x(i) with subsequent columns y(-N),y(-N+1)...y(0),y(1),...y(N)

Typically the negative-ranging indices might correspond to ordered gridpoints along a spatial axis "y", scattered on either size of y=0.

At present the negative numbers (intended to serve as column identifiers) are interpreted as pertaining to provision of error data. Is there an easy way around this? Thanks, JohnW

Anonymous

 

(Note: upload size limit is set to 1024 kB, after insertion of the required escape characters.)

Attach File(s):
   
   
Comment:
   

Attached Files
file #12747:  rampCKLS.vsz added by jeremysanders (563B - application/octet-stream)
file #12745:  rampCKLS.dat added by jdwilson (130kB - video/mpeg - Yet another version of the data file, this time with a letter before the minus signs)
file #12746:  whatveuszsees.dat added by jdwilson (1kB - video/mpeg - Yet another version of the data file, this time with a letter before the minus signs)
file #12742:  rampCKLS.dat added by jdwilson (129kB - video/mpeg)
file #12744:  rampCKLS_generic.dat added by jdwilson (129kB - video/mpeg)

 

Depends on the following items: None found

Items that depend on this one: None found

 

Carbon-Copy List
  • -unavailable- added by jdwilson (Posted a comment)
  • -unavailable- added by jeremysanders (Posted a comment)
  • -unavailable- added by None (Submitted the item)
  •  

    Do you think this task is very important?
    If so, you can click here to add your encouragement to it.
    This task has 0 encouragements so far.

    Only logged-in users can vote.

     

    Please enter the title of George Orwell's famous dystopian book (it's a date):

     

     

    Follow 6 latest changes.

    Date Changed By Updated Field Previous Value => Replaced By
    Wed 30 Mar 2011 08:30:00 PM UTCjeremysandersAttached File-=>Added rampCKLS.vsz, #12747
    Wed 30 Mar 2011 08:04:11 PM UTCjdwilsonAttached File-=>Added rampCKLS.dat, #12745
      Attached File-=>Added whatveuszsees.dat, #12746
    Wed 30 Mar 2011 06:50:00 PM UTCjdwilsonAttached File-=>Added rampCKLS.dat, #12742
      Attached File-=>Added rampCKLS_withbackquotes.dat, #12743
      Attached File-=>Added rampCKLS_generic.dat, #12744
    Show feedback again

    Back to the top


    Powered by Savane 3.1-cleanup