I suspect it's an encoding problem

Ruud
http://svg.startpagina.nl

Hi

If I plot using pl's internal x11 interface
pl test.plotdefs -font "Arial" unicode works correct

I use the following:
pl test.plotdefs -font "Arial" -png -o stdout > test.png

Unicode characters are not printed correctly.
It does not seem to honour the -font setting either

Is there any other way to print to stdout which work like the x11
interface.

Is it a limitation in the console (it uses an unicode font)

I use various linux suse10, redhat4
pl version: ploticus 2.33-Jun'06 (unix)

Thanks for a great program

--
Morten Bjoernsvik, Oslo, Norway

When i use the below code, my first boxplot is complete but my second is only
half
complete. The min, mean values are not shown.

Whats wrong with the code ? I can't find it.

Thanks in advance
Fred

#proc getdata
data:
24988 24779
24832 24779
24832 24779
24832 24779
24073 24073
24073 24073
24073 24073
24073 24073
24073 24073
24073 24073
24173 24073
24173 24073
24073 24073
24173 24073
24173 24073
24173 24073
24073 24073
24073 24073
23916 24073
23916 24073
24073 24073
24073 24073
24073 24073
24073 24073
24073 24073
24073 24073
24073 24073
24073 24073
23916 24073
24073 24073
24073 24073
24073 24073
24073 24073
24073 24073
24073 23916
24073 23916
24073 24073
24073 24073
24073 24073
24073 24073
24073 24073
24073 24073
24073 24073
24073 24073
24073 24073
24073 24073
24073 24073
24073 24073
24073 24073
23916 24073
24073 24073
24073 24073
24073 24073
24073 24073
24073 24073
24073 24073
24073 24073
24073 24073
24073 24073
24073 24073


#proc areadef
   title: NL6
   rectangle: 1 1 5 3
   xrange: 0 3
   yrange: 23500 25000
   axes: none

#proc yaxis
   grid: color=black
   stubs: none
   tics: none
   axisline: none

#proc rangebar
   datafield: 1
   barloc: 1
   showstats: yes
   color: white
   mediansym: radius=0
   meansym: shape=circle style=fill fillcolor=black radius=0.04

#proc rangebar
   datafield: 2
   barloc: 2
   showstats: yes
   color: white
   mediansym: radius=0
   meansym: shape=circle style=fill fillcolor=blue radius=0.04

Ouput :

// Rangebar statistics computed on data field 1
// N=60,  mean=24124.1,  stddev=208.204,  #missing=0,  sum(E)=1.44744e+06
// min=23916,  max=24988,  median=24073,  25th_pctile=24073,  75th_pctile=24073,
IQR=0
// Tails (mode=5/95): low=23916,  high=24832

// Rangebar statistics computed on data field 2
// N=60,  mean=24114.8,  stddev=181.241,  #missing=0,  sum(E)=1.44689e+06
// min=23916,  max=24779,  median=24073,  25th_pctile=24073,  75th_pctile=24073,
IQR=0
// Tails (mode=5/95): low=24073,  high=24779

Hi Steve,

It seems like dates before 1754 cannot be plotted, but instead start
plotting at 1970.  Is there a workaround for that?

Dmitry think it might be to do with this:
http://www.litencyc.com/php/stopics.php?rec=true&UID=141

And the code seems to confirm that, from dates.c:
#define BASEYR  1970  /* 1970 = start of unix epoch */
#define BASEOLDYR  1754  /* old, but not so old as to run into Sep
1752 anomoly */
...
if( y < BASEYR ) refyear = BASEOLDYR;  /* switch to BASEOLDYR as year
basis for calculations */

Brian, perhaps you could simply plot years as integers, and use plain
numeric scale?   On Sept 14, 1752 the British empire adopted the Gregorian
calendar and skipped 11 days (Sept 2 was immediately followed by Sept 14).

-Steve

On Sat, 20 Jan 2007, Brian Mulloy wrote:

> Hi Steve,
>
> It seems like dates before 1754 cannot be plotted, but instead start
> plotting at 1970.  Is there a workaround for that?
>
> Dmitry think it might be to do with this:
> http://www.litencyc.com/php/stopics.php?rec=true&UID=141
>
> And the code seems to confirm that, from dates.c:
> #define BASEYR  1970  /* 1970 = start of unix epoch */
> #define BASEOLDYR  1754  /* old, but not so old as to run into Sep
> 1752 anomoly */
> ...
> if( y < BASEYR ) refyear = BASEOLDYR;  /* switch to BASEOLDYR as year
> basis for calculations */
>
>
>


   Stephen C. Grubb     scg@...                x-6633
   Scientific Software Engineer, The Jackson Laboratory
   600 Main Street  Bar Harbor, Maine 04609 USA

I'm having trouble with cumufrac and can't figure out what's wrong. Hoping
someone
can help. I'm using Ploticus 2.32 on Mac OS X 10.4.8 (prebuilt from the Ploticus
web
site.)

Briefly, the first command below creates a correct cumufrac gif file. The second
command creates a gif with the correct axes extents, but no data line--the
plotting
area is blank.

works:           bin/pl -debug -gif -prefab cumufrac x=2 data=foo echodata=yes
doesn't work: bin/pl -debug -gif -prefab cumufrac x=1 data=foo echodata=yes

foo contains:

1       1
2       2
3       3
4       4
5       5
6       10
7       23
8       2

debug output for the working example is:

bin/pl -debug -gif -prefab cumufrac x=2 data=foo echodata=yes

Got command line arg(s): -gif
Prefabs dir is: /Users/jsimons/Desktop/Downloads/pl232osx/prefabs
Script file is /Users/jsimons/Desktop/Downloads/pl232osx/prefabs/cumufrac.pl
Got command line arg(s): x=2
Got command line arg(s): data=foo
Got command line arg(s): echodata=yes
Device code is g
Setting output file name to cumufrac.gif
Version: pl 2.32
Script file is: /Users/jsimons/Desktop/Downloads/pl232osx/prefabs/cumufrac.pl
Script file successfully opened
Executing settings
Executing getdata
filling data set# 0
getdata has read 8 records; there are 2 fields per record.
// proc getdata has read & parsed these data:
[1][1]
[2][2]
[3][3]
[4][4]
[5][5]
[6][10]
[7][23]
[8][2]
Got 8 records, 2 fields per record.
(endproc)
Executing areadef
Autorange on x: min=0 to max=25
areadef: lowerleft: 2,2  upperright: 7,5
areadef:   xrange is 0 to 25.   yrange is 0 to 8.
(endproc)
Executing xaxis
Executing lineplot
sorting points for line
Executing line
Executing areadef
areadef: lowerleft: 2,2  upperright: 7,5
areadef:   xrange is 0 to 1.   yrange is 0 to 1.
Done with page.  Writing out result file.  Computed bounding box is: 1.43 , 1.59
to 7.20 ,
5.20

debug output for the non-working example is:


Got command line arg(s): -gif
Prefabs dir is: /Users/jsimons/Desktop/Downloads/pl232osx/prefabs
Script file is /Users/jsimons/Desktop/Downloads/pl232osx/prefabs/cumufrac.pl
Got command line arg(s): x=1
Got command line arg(s): data=foo
Got command line arg(s): echodata=yes
Device code is g
Setting output file name to cumufrac.gif
Version: pl 2.32
Script file is: /Users/jsimons/Desktop/Downloads/pl232osx/prefabs/cumufrac.pl
Script file successfully opened
Executing settings
Executing getdata
filling data set# 0
getdata has read 8 records; there are 2 fields per record.
// proc getdata has read & parsed these data:
[1][1]
[2][2]
[3][3]
[4][4]
[5][5]
[6][10]
[7][23]
[8][2]
Got 8 records, 2 fields per record.
(endproc)
Executing areadef
Autorange on x: min=1 to max=9
areadef: lowerleft: 2,2  upperright: 7,5
areadef:   xrange is 1 to 9.   yrange is 0 to 8.
(endproc)
Executing xaxis
Executing lineplot
sorting points for line
Executing line
Executing areadef
areadef: lowerleft: 2,2  upperright: 7,5
areadef:   xrange is 0 to 1.   yrange is 0 to 1.
Done with page.  Writing out result file.  Computed bounding box is: 1.43 , 1.59
to 7.20 ,
5.20

Having worked more with Ploticus since I originally sent in my question about
cumufrac
(I am working on a project to convince my town that they are not assessing our
neighborhood
fairly), the reported problem may be a symptom of either a larger problem or
just my
lack of understanding.

I find that for some combinations of barsize and binsize, I get no data in my
graph. I
haven't been able to understand what is going on at all--I can't predict which
values
will result in a data display and which will cause a blank display. Again, it
may be
me or maybe there is a problem.

In the original problem report, I didn't specify either binsize or barsize so I
would have
expected Ploticus to pick appropriate values. That the display is blank leads me
to
wonder if there really is a problem beyond my lack of understanding.

Any pointers or help would be much appreciated.

Josh Simons

Hi, thanks for the very clear problem report.. I wish they were all passed
to me this way.

It looks like a bug, am continuing to look into it.

I'll get back to you soon,

Steve


On Sat, 10 Feb 2007, ibid17 wrote:

>
> I'm having trouble with cumufrac and can't figure out what's wrong. Hoping
someone
> can help. I'm using Ploticus 2.32 on Mac OS X 10.4.8 (prebuilt from the
Ploticus web
> site.)
>
> Briefly, the first command below creates a correct cumufrac gif file. The
second
> command creates a gif with the correct axes extents, but no data line--the
plotting
> area is blank.
>
> works:           bin/pl -debug -gif -prefab cumufrac x=2 data=foo echodata=yes
> doesn't work: bin/pl -debug -gif -prefab cumufrac x=1 data=foo echodata=yes
>
> foo contains:
>
> 1       1
> 2       2
> 3       3
> 4       4
> 5       5
> 6       10
> 7       23
> 8       2
>
> debug output for the working example is:
>
> bin/pl -debug -gif -prefab cumufrac x=2 data=foo echodata=yes
>
> Got command line arg(s): -gif
> Prefabs dir is: /Users/jsimons/Desktop/Downloads/pl232osx/prefabs
> Script file is /Users/jsimons/Desktop/Downloads/pl232osx/prefabs/cumufrac.pl
> Got command line arg(s): x=2
> Got command line arg(s): data=foo
> Got command line arg(s): echodata=yes
> Device code is g
> Setting output file name to cumufrac.gif
> Version: pl 2.32
> Script file is: /Users/jsimons/Desktop/Downloads/pl232osx/prefabs/cumufrac.pl
> Script file successfully opened
> Executing settings
> Executing getdata
> filling data set# 0
> getdata has read 8 records; there are 2 fields per record.
> // proc getdata has read & parsed these data:
> [1][1]
> [2][2]
> [3][3]
> [4][4]
> [5][5]
> [6][10]
> [7][23]
> [8][2]
> Got 8 records, 2 fields per record.
> (endproc)
> Executing areadef
> Autorange on x: min=0 to max=25
> areadef: lowerleft: 2,2  upperright: 7,5
> areadef:   xrange is 0 to 25.   yrange is 0 to 8.
> (endproc)
> Executing xaxis
> Executing lineplot
> sorting points for line
> Executing line
> Executing areadef
> areadef: lowerleft: 2,2  upperright: 7,5
> areadef:   xrange is 0 to 1.   yrange is 0 to 1.
> Done with page.  Writing out result file.  Computed bounding box is: 1.43 ,
1.59  to 7.20 ,
> 5.20
>
> debug output for the non-working example is:
>
>
> Got command line arg(s): -gif
> Prefabs dir is: /Users/jsimons/Desktop/Downloads/pl232osx/prefabs
> Script file is /Users/jsimons/Desktop/Downloads/pl232osx/prefabs/cumufrac.pl
> Got command line arg(s): x=1
> Got command line arg(s): data=foo
> Got command line arg(s): echodata=yes
> Device code is g
> Setting output file name to cumufrac.gif
> Version: pl 2.32
> Script file is: /Users/jsimons/Desktop/Downloads/pl232osx/prefabs/cumufrac.pl
> Script file successfully opened
> Executing settings
> Executing getdata
> filling data set# 0
> getdata has read 8 records; there are 2 fields per record.
> // proc getdata has read & parsed these data:
> [1][1]
> [2][2]
> [3][3]
> [4][4]
> [5][5]
> [6][10]
> [7][23]
> [8][2]
> Got 8 records, 2 fields per record.
> (endproc)
> Executing areadef
> Autorange on x: min=1 to max=9
> areadef: lowerleft: 2,2  upperright: 7,5
> areadef:   xrange is 1 to 9.   yrange is 0 to 8.
> (endproc)
> Executing xaxis
> Executing lineplot
> sorting points for line
> Executing line
> Executing areadef
> areadef: lowerleft: 2,2  upperright: 7,5
> areadef:   xrange is 0 to 1.   yrange is 0 to 1.
> Done with page.  Writing out result file.  Computed bounding box is: 1.43 ,
1.59  to 7.20 ,
> 5.20
>
>


   Stephen C. Grubb     scg@...                x-6633
   Scientific Software Engineer, The Jackson Laboratory
   600 Main Street  Bar Harbor, Maine 04609 USA

Hi,

attached is a revised copy of 'cumufrac.pl' ... place it in your ploticus
prefabs directory.

this fixes a bug recently reported with the 'cumufrac' prefab, where no
line was being drawn for certain input data.  In short, I added
'mininit=0' parameter to the xautorange statement.  The bug was being
caused by automatic plot range in x not having a fixed 0 minima, so for
some data the minima was non-zero, which in turn caused line drawing to
never get started.  For more info or to confirm, see previous message
below.

Let me know if any further difficulty is encountered related to this.
Thanks,

Steve Grubb


---------- Forwarded message ----------
Date: Tue, 13 Feb 2007 10:19:44 -0500 (EST)
From: Stephen C. Grubb <scg@...>
Reply-To: ploticus@yahoogroups.com
To: ploticus@yahoogroups.com
Subject: Re: [ploticus] Cumufrac problem

Hi, thanks for the very clear problem report.. I wish they were all passed
to me this way.

It looks like a bug, am continuing to look into it.

I'll get back to you soon,

Steve


On Sat, 10 Feb 2007, ibid17 wrote:

>
> I'm having trouble with cumufrac and can't figure out what's wrong. Hoping
someone
> can help. I'm using Ploticus 2.32 on Mac OS X 10.4.8 (prebuilt from the
Ploticus web
> site.)
>
> Briefly, the first command below creates a correct cumufrac gif file. The
second
> command creates a gif with the correct axes extents, but no data line--the
plotting
> area is blank.
>
> works:           bin/pl -debug -gif -prefab cumufrac x=2 data=foo echodata=yes
> doesn't work: bin/pl -debug -gif -prefab cumufrac x=1 data=foo echodata=yes
>
> foo contains:
>
> 1       1
> 2       2
> 3       3
> 4       4
> 5       5
> 6       10
> 7       23
> 8       2
>
> debug output for the working example is:
>
> bin/pl -debug -gif -prefab cumufrac x=2 data=foo echodata=yes
>
> Got command line arg(s): -gif
> Prefabs dir is: /Users/jsimons/Desktop/Downloads/pl232osx/prefabs
> Script file is /Users/jsimons/Desktop/Downloads/pl232osx/prefabs/cumufrac.pl
> Got command line arg(s): x=2
> Got command line arg(s): data=foo
> Got command line arg(s): echodata=yes
> Device code is g
> Setting output file name to cumufrac.gif
> Version: pl 2.32
> Script file is: /Users/jsimons/Desktop/Downloads/pl232osx/prefabs/cumufrac.pl
> Script file successfully opened
> Executing settings
> Executing getdata
> filling data set# 0
> getdata has read 8 records; there are 2 fields per record.
> // proc getdata has read & parsed these data:
> [1][1]
> [2][2]
> [3][3]
> [4][4]
> [5][5]
> [6][10]
> [7][23]
> [8][2]
> Got 8 records, 2 fields per record.
> (endproc)
> Executing areadef
> Autorange on x: min=0 to max=25
> areadef: lowerleft: 2,2  upperright: 7,5
> areadef:   xrange is 0 to 25.   yrange is 0 to 8.
> (endproc)
> Executing xaxis
> Executing lineplot
> sorting points for line
> Executing line
> Executing areadef
> areadef: lowerleft: 2,2  upperright: 7,5
> areadef:   xrange is 0 to 1.   yrange is 0 to 1.
> Done with page.  Writing out result file.  Computed bounding box is: 1.43 ,
1.59  to 7.20 ,
> 5.20
>
> debug output for the non-working example is:
>
>
> Got command line arg(s): -gif
> Prefabs dir is: /Users/jsimons/Desktop/Downloads/pl232osx/prefabs
> Script file is /Users/jsimons/Desktop/Downloads/pl232osx/prefabs/cumufrac.pl
> Got command line arg(s): x=1
> Got command line arg(s): data=foo
> Got command line arg(s): echodata=yes
> Device code is g
> Setting output file name to cumufrac.gif
> Version: pl 2.32
> Script file is: /Users/jsimons/Desktop/Downloads/pl232osx/prefabs/cumufrac.pl
> Script file successfully opened
> Executing settings
> Executing getdata
> filling data set# 0
> getdata has read 8 records; there are 2 fields per record.
> // proc getdata has read & parsed these data:
> [1][1]
> [2][2]
> [3][3]
> [4][4]
> [5][5]
> [6][10]
> [7][23]
> [8][2]
> Got 8 records, 2 fields per record.
> (endproc)
> Executing areadef
> Autorange on x: min=1 to max=9
> areadef: lowerleft: 2,2  upperright: 7,5
> areadef:   xrange is 1 to 9.   yrange is 0 to 8.
> (endproc)
> Executing xaxis
> Executing lineplot
> sorting points for line
> Executing line
> Executing areadef
> areadef: lowerleft: 2,2  upperright: 7,5
> areadef:   xrange is 0 to 1.   yrange is 0 to 1.
> Done with page.  Writing out result file.  Computed bounding box is: 1.43 ,
1.59  to 7.20 ,
> 5.20
>
>


   Stephen C. Grubb     scg@...                x-6633
   Scientific Software Engineer, The Jackson Laboratory
   600 Main Street  Bar Harbor, Maine 04609 USA

// ploticus data display engine.  Software, documentation, and examples.
// Copyright 1998-2002 Stephen C. Grubb  (scg@...).
// Covered by GPL; see the file 'Copyright' for details.
// http://ploticus.sourceforge.net

// CUMUFRAC - cumulative fraction plot

// logmode 	 pass as 'log' for log plot
// x x2
// linedet, linedet2
// name, name2


#setifnotgiven logmode = linear
#setifnotgiven x2 = ""
#setifnotgiven rectangle = "2 2 7 5"
     // (what about cm units??)
#setifnotgiven linedet = "color=green width=0.5"
#setifnotgiven linedet2 = "color=red width=0.5"

#setifnotgiven name = "#usefname"
#setifnotgiven name2 = "#usefname"
#if @CM_UNITS = 1
   #setifnotgiven legend = "min+1.2 min-1.2"
#else
   #setifnotgiven legend = "min+0.5 min-0.5"
#endif


/// get standard parameters..
#include $chunk_setstd

/// read the data for the 1st curve..
/// the select screens out missing data codes and other non-numerics..
#set select = "$isnumber(@" @x ") = 1"
#include $chunk_read


#musthave x

// chunk_area uses unittyp...
#set unittype = @logmode

/// set up plotting area...
#include $chunk_area
yrange: 0 @NRECORDS
#if @logmode = linear
   // xautorange: datafield=@x,@x2 nearest=@xnearest  // changed as below, scg
2/20/07...
   xautorange: datafield=@x,@x2 mininit=0 nearest=@xnearest
#else
   xautorange: datafield=@x,@x2 mininit=0.05 nearest=@xnearest
#endif
#endproc



///// do x axis (y axis done below)..
#proc xaxis
#ifspec xlbl label
#ifspec xlbldet labeldetails
#if @logmode != "linear"
   selflocatingstubs: text
   #include $chunk_logstubs
#else
   stubs: inc @xinc
#endif
#ifspec @stubfmt
stubrange: @XMIN
#ifspec @stubvert
#ifspec xgrid grid
location: min-0.2
stubcull: yes


#if @logmode != "linear"
   // do log tic marks..
   #proc xaxis
   location: min-0.2
   selflocatingstubs: text
   #include $chunk_logtics
#endif

/// title..
#include $chunk_title

/// user pre-plot include...
#if @include1 != ""
   #include @include1
#endif

///// plot the curve using accum and instancemode..
#proc lineplot
xfield: @x
linedetails: @linedet
sort: yes
instancemode: yes
accum: yes
stairstep: yes
#if @logmode = "linear"
   firstpoint: 0 0
#endif
lastseglen: 0.2
legendlabel: @name
#saveas L

#if @x2 != ""
   /// read the data for the 2nd curve..
   /// the select screens out missing data codes and other non-numerics..
   #set select = "$isnumber(@" @x2 ") = 1"
   #include $chunk_read

   #include $chunk_area
   yrange: 0 @NRECORDS
   xrange: @XMIN @XMAX
       // set in first areadef above..

   // plot 2nd curve
   #proc lineplot
   #clone: L
   xfield: @x2
   linedetails: @linedet2
   legendlabel: @name2
#endif


///// draw a plotting area frame allowing a small margin below X=0 and below
Y=0..
#proc line
linedetails: width=0.5 color=black
points: min-0.2 min-0.2 max+0.2 min-0.2
	 max+0.2 max+0.2
	 min-0.2 max+0.2
	 min-0.2 min-0.2


///// redraw the plotting area so we can express Y axis as a percentage 0.0 to
1.0 ..
#proc areadef
rectangle: @rectangle
xrange: 0 1
yrange: 0 1
yaxis.stubs: inc 0.2
yaxis.location: min-0.2



// do legend
#if @legend != "no"
   #if @name != "#usefname" || @header = yes
     #proc legend
     location: @legend
     #ifspec legendfmt format
     #ifspec legendsep sep
   #ifspec legwrap wraplen
   #ifspec legbreak extent
   #ifspec legtitle title
   #ifspec legbox backcolor
   #ifspec legframe frame
   #ifspec legtextdet textdetails

   #endif
#endif


//// user post-plot include..
#if @include2 != ""
   #include @include2
#endif

Fred,  I finally got a chance to check into this problem.

I believe what you're seeing is caused by the characteristics of the data,
and because by default the boxplot tails are at the 5th and 95th
percentile.  In your second case, the median (shown by a lighter hash
mark) is the exact same value as the 5th percentile.  Hence the median
only is shown.

If you want the boxplot tails to show min and max you should add this
statement:

proc rangebar
...
tailmode: minmax


Thanks,

Steve


On Wed, 27 Dec 2006, frederick.ryckbosch wrote:

> When i use the below code, my first boxplot is complete but my second is only
half
> complete. The min, mean values are not shown.
>
> Whats wrong with the code ? I can't find it.
>
> Thanks in advance
> Fred
>
> #proc getdata
> data:
> 24988 24779
> 24832 24779
> 24832 24779
> 24832 24779
> 24073 24073
> 24073 24073
> 24073 24073
> 24073 24073
> 24073 24073
> 24073 24073
> 24173 24073
> 24173 24073
> 24073 24073
> 24173 24073
> 24173 24073
> 24173 24073
> 24073 24073
> 24073 24073
> 23916 24073
> 23916 24073
> 24073 24073
> 24073 24073
> 24073 24073
> 24073 24073
> 24073 24073
> 24073 24073
> 24073 24073
> 24073 24073
> 23916 24073
> 24073 24073
> 24073 24073
> 24073 24073
> 24073 24073
> 24073 24073
> 24073 23916
> 24073 23916
> 24073 24073
> 24073 24073
> 24073 24073
> 24073 24073
> 24073 24073
> 24073 24073
> 24073 24073
> 24073 24073
> 24073 24073
> 24073 24073
> 24073 24073
> 24073 24073
> 24073 24073
> 23916 24073
> 24073 24073
> 24073 24073
> 24073 24073
> 24073 24073
> 24073 24073
> 24073 24073
> 24073 24073
> 24073 24073
> 24073 24073
> 24073 24073
>
>
> #proc areadef
>   title: NL6
>   rectangle: 1 1 5 3
>   xrange: 0 3
>   yrange: 23500 25000
>   axes: none
>
> #proc yaxis
>   grid: color=black
>   stubs: none
>   tics: none
>   axisline: none
>
> #proc rangebar
>   datafield: 1
>   barloc: 1
>   showstats: yes
>   color: white
>   mediansym: radius=0
>   meansym: shape=circle style=fill fillcolor=black radius=0.04
>
> #proc rangebar
>   datafield: 2
>   barloc: 2
>   showstats: yes
>   color: white
>   mediansym: radius=0
>   meansym: shape=circle style=fill fillcolor=blue radius=0.04
>
> Ouput :
>
> // Rangebar statistics computed on data field 1
> // N=60,  mean=24124.1,  stddev=208.204,  #missing=0,  sum(E)=1.44744e+06
> // min=23916,  max=24988,  median=24073,  25th_pctile=24073, 
75th_pctile=24073,
> IQR=0
> // Tails (mode=5/95): low=23916,  high=24832
>
> // Rangebar statistics computed on data field 2
> // N=60,  mean=24114.8,  stddev=181.241,  #missing=0,  sum(E)=1.44689e+06
> // min=23916,  max=24779,  median=24073,  25th_pctile=24073, 
75th_pctile=24073,
> IQR=0
> // Tails (mode=5/95): low=24073,  high=24779
>
>
>


   Stephen C. Grubb     scg@...                x-6633
   Scientific Software Engineer, The Jackson Laboratory
   600 Main Street  Bar Harbor, Maine 04609 USA

Hi,

I've confirmed that ploticus 2.33 generates an svg where clicking on data
points works correctly (when tried on IE).

My 2.33 includes the 8 Sep 06 patch related to svg encoding
http://ploticus.sourceforge.net/doc/news.html

I'll fix the web site asap, but currently sourceforge shell access is down
for unknown reasons.

Thanks,

Steve

On Fri, 29 Dec 2006, steltenpower wrote:

> I suspect it's an encoding problem
>
> Ruud
> http://svg.startpagina.nl
>
>


   Stephen C. Grubb     scg@...                x-6633
   Scientific Software Engineer, The Jackson Laboratory
   600 Main Street  Bar Harbor, Maine 04609 USA

Hi Morten,

One option would be to build ploticus with freetype2 font support.. see
the Makefile.  This would allow generation of PNG to stdout, using
freetype2 fonts, and using a freetype2-aware version of GD.

I don't really know what else to tell you, besides what is described on
the fonts man page http://ploticus.sourceforge.net/doc/fonts.html

Let me know how this goes,

-Steve


On Tue, 2 Jan 2007, [iso-8859-1] Morten Bjørnsvik wrote:

> Hi
>
> If I plot using pl's internal x11 interface
> pl test.plotdefs -font "Arial" unicode works correct
>
> I use the following:
> pl test.plotdefs -font "Arial" -png -o stdout > test.png
>
> Unicode characters are not printed correctly.
> It does not seem to honour the -font setting either
>
> Is there any other way to print to stdout which work like the x11
> interface.
>
> Is it a limitation in the console (it uses an unicode font)
>
> I use various linux suse10, redhat4
> pl version: ploticus 2.33-Jun'06 (unix)
>
> Thanks for a great program
>
> --
> Morten Bjoernsvik, Oslo, Norway
>
>


   Stephen C. Grubb     scg@...                x-6633
   Scientific Software Engineer, The Jackson Laboratory
   600 Main Street  Bar Harbor, Maine 04609 USA

Hi,

I forgot about linewrapping in email message body which may cause problems
with the earlier message(s) containing revised cumufrac.pl code.

Another way to execute this fix would simply be to go to your ploticus
prefabs directory, edit cumufrac.pl, and change line 48 so it looks like
line 49 below.



44  /// set up plotting area...
45  #include $chunk_area
46  yrange: 0 @NRECORDS
47  #if @logmode = linear
48    // xautorange: datafield=@x,@x2 nearest=@xnearest
49    xautorange: datafield=@x,@x2 mininit=0 nearest=@xnearest
50  #else
51    xautorange: datafield=@x,@x2 mininit=0.05 nearest=@xnearest
52  #endif
53  #endproc


Thanks,

Steve

Hello. I'm a rockie with ploticus and I'm trying examples from
prefabs. When I do the chron example 8 - multiple curves I can only
get one curve . No error messages shown.

What is wrong? Please help! Thanks!

FC6 with:

ploticus 2.33-Jun'06 (unix)
Linux dhcppc0 2.6.19-1.2911.fc6 #1 SMP Sat Feb 10 15:51:47 EST 2007
i686 athlon i386 GNU/Linux

[micko@dhcppc0 ~]$ echo $PLOTICUS_PREFABS
/usr/local/src/plsrc200/prefabs

pl -prefab chron data=test.txt header=yes unittype=datetime mode=line
datefmt=mm/dd/yyyy x=dt y=airtemp y2=soiltemp y3=watertemp   xinc="1
day" legendfmt=singleline


dt airtemp soiltemp watertemp
12/08/2002.06:00 24 27 16
12/09/2002.00:00 25 27 16
12/10/2002.00:00 30 27 15
12/11/2002.00:00 32 32 15
12/12/2002.00:00 40 37 16
12/13/2002.00:00 19 27 15
12/14/2002.00:00 14 18 13

Hi:

How can I change the number of pixels between each stub?

Hello,
We are running ploticus 1.41 on Solaris 10 stations.
It used to work ok but recently we met some bugs.
Maybe because the amount of data has increased or the duration of the
event covered in the graph... I don't know.

To check if we can avoid the problem with version 2.0 I'm looking for
the source that I did not find on the Web neither on the ploticus
website. I tried version 2.33 but two many procedures and functions
have changed between the two versions...

Thanks for your help.
Eric

Greetings,

uname -a results for OS: Linux ted1 2.6.15-1.2054_FC5smp #1 SMP Tue
Mar 14 16:05:46 EST 2006 i686 i686 i386 GNU/Linux

Hopefully I can get this program to work as I am excited to use it.
I downloaded the precompiled binaries and tried to make a stack graph
and the processing seems to be fine but my result graph is just a
black box.

I then downloaded the src and compiled it myself, retried with same
result.

I then tried the prefab stack example.

debug.txt:
0 100 2 100 4 100 3 100 4
1.5 92 4 94 7 87 6 86 5
3 89 4 92 8 78 6 74 4

command line:
[root@ted1 src]# pl -prefab stack data=/root/graphdata/debug.txt x=1
y=2 y2=4 y3=6 y4=8 -debug

Device code is x
Version: pl 2.33-Jun'06
Script file is: /root/ploticus/pl233src/prefabs/stack.pl
Script file successfully opened
Executing settings
Executing getdata
filling data set# 0
getdata has read 3 records; there are 9 fields per record.
Got 3 records, 9 fields per record.
(endproc)
Executing areadef
categories in x: list of size=250 malloced
Autorange on y: min=0 to max=500
areadef: lowerleft: 1,2  upperright: 5.5,4
areadef:   xrange is 0 to 4.   yrange is 0 to 500.
Executing xaxis
Executing yaxis
Executing bars
Executing bars
Executing bars
Executing bars

but still get the black box for a result.

I then tried loading the stock.htm and got the same graph result but
with a few  note comments added:
[root@ted1 src]# pl stock.htm -debug

Device code is x
Version: pl 2.33-Jun'06
Script file is: stock.htm
Script file successfully opened
Executing page
Executing getdata
filling data set# 0
getdata has read 92 records; there are 6 fields per record.
Executing transform
filling data set# 1 (this will now be the current data set)
Executing areadef
areadef: lowerleft: 1,3  upperright: 5,5
areadef:   xrange is 01-Jan-99 to 01-Jun-99.   yrange is 45 to 65.
Executing xaxis
Executing yaxis
Executing bars
note: pl proc bars encountered 1 unplottable data values
Executing lineplot
note: pl proc lineplot encountered 1 unplottable data values
Executing curvefit
note: pl proc curvefit encountered 1 unplottable data values
Executing legend
Executing areadef
areadef: lowerleft: 1,1.6  upperright: 5,2.6
areadef:   xrange is 01-Jan-99 to 01-Jun-99.   yrange is 0 to 30000000.
Executing xaxis
Executing yaxis
Executing bars
note: pl proc bars encountered 1 unplottable data values

Hopefully, this is a simple config issue and any advice or suggestions
would be greatly appreciated.

Regards,
~L

Hi,

it seems like it might be an x11-related problem.. perhaps theres an xlib
version issue or something like that.  Is that a fairly old linux version?

Another thing to try, to verify that the pl engine itself is working ok...
are you able to create a png image?

(pl -png -prefab stat .....)


-Steve


On Wed, 14 Mar 2007, Actaeus wrote:

> Greetings,
>
> uname -a results for OS: Linux ted1 2.6.15-1.2054_FC5smp #1 SMP Tue
> Mar 14 16:05:46 EST 2006 i686 i686 i386 GNU/Linux
>
> Hopefully I can get this program to work as I am excited to use it.
> I downloaded the precompiled binaries and tried to make a stack graph
> and the processing seems to be fine but my result graph is just a
> black box.
>
> I then downloaded the src and compiled it myself, retried with same
> result.
>
> I then tried the prefab stack example.
>
> debug.txt:
> 0 100 2 100 4 100 3 100 4
> 1.5 92 4 94 7 87 6 86 5
> 3 89 4 92 8 78 6 74 4
>
> command line:
> [root@ted1 src]# pl -prefab stack data=/root/graphdata/debug.txt x=1
> y=2 y2=4 y3=6 y4=8 -debug
>
> Device code is x
> Version: pl 2.33-Jun'06
> Script file is: /root/ploticus/pl233src/prefabs/stack.pl
> Script file successfully opened
> Executing settings
> Executing getdata
> filling data set# 0
> getdata has read 3 records; there are 9 fields per record.
> Got 3 records, 9 fields per record.
> (endproc)
> Executing areadef
> categories in x: list of size=250 malloced
> Autorange on y: min=0 to max=500
> areadef: lowerleft: 1,2  upperright: 5.5,4
> areadef:   xrange is 0 to 4.   yrange is 0 to 500.
> Executing xaxis
> Executing yaxis
> Executing bars
> Executing bars
> Executing bars
> Executing bars
>
> but still get the black box for a result.
>
> I then tried loading the stock.htm and got the same graph result but
> with a few  note comments added:
> [root@ted1 src]# pl stock.htm -debug
>
> Device code is x
> Version: pl 2.33-Jun'06
> Script file is: stock.htm
> Script file successfully opened
> Executing page
> Executing getdata
> filling data set# 0
> getdata has read 92 records; there are 6 fields per record.
> Executing transform
> filling data set# 1 (this will now be the current data set)
> Executing areadef
> areadef: lowerleft: 1,3  upperright: 5,5
> areadef:   xrange is 01-Jan-99 to 01-Jun-99.   yrange is 45 to 65.
> Executing xaxis
> Executing yaxis
> Executing bars
> note: pl proc bars encountered 1 unplottable data values
> Executing lineplot
> note: pl proc lineplot encountered 1 unplottable data values
> Executing curvefit
> note: pl proc curvefit encountered 1 unplottable data values
> Executing legend
> Executing areadef
> areadef: lowerleft: 1,1.6  upperright: 5,2.6
> areadef:   xrange is 01-Jan-99 to 01-Jun-99.   yrange is 0 to 30000000.
> Executing xaxis
> Executing yaxis
> Executing bars
> note: pl proc bars encountered 1 unplottable data values
>
> Hopefully, this is a simple config issue and any advice or suggestions
> would be greatly appreciated.
>
> Regards,
> ~L
>
>


   Stephen C. Grubb     scg@...                x-6633
   Scientific Software Engineer, The Jackson Laboratory
   600 Main Street  Bar Harbor, Maine 04609 USA

Hello.

I am trying to graph hits per hour using the chron prefab. I am
running into trouble with specifying an xrange. My dataset spans
severals days over several weeks. I want to graph all hits in a single
day specified by my xrange= . Ploticus seems to try and graph every
day present in my dataset instead of just the day I specified with
xrange=. Anyone have any clues?

I amm calling ploticus as follows

pl -backcolor black  -color white  -png -o @ARGV[1] -prefab chron
data=@ARGV[0] xrange= 22-mar-2007.0:0:0 22-mar-2007.23:59:59
tab=hour  datefmt=dd/mmm/yyyy xmargin=0  xinc=\"1 hour\" barwidth=line
  stubfmt=hh unittype=datetime timefld=2  title=\"\# hits per hour\"
color=green

my dataset is as follows:

19/Mar/2007 23:04:36
19/Mar/2007 23:05:27
19/Mar/2007 23:08:27
19/Mar/2007 23:08:40
19/Mar/2007 23:11:53
19/Mar/2007 23:11:58
19/Mar/2007 23:16:55
20/Mar/2007 01:10:10
20/Mar/2007 05:40:53
20/Mar/2007 05:43:57
20/Mar/2007 12:40:25
20/Mar/2007 19:22:04
20/Mar/2007 21:14:49
21/Mar/2007 01:26:09
21/Mar/2007 09:29:03
21/Mar/2007 14:26:34
21/Mar/2007 18:10:08
22/Mar/2007 03:01:19
22/Mar/2007 11:13:51
22/Mar/2007 11:19:22
22/Mar/2007 14:12:15
22/Mar/2007 19:45:46
22/Mar/2007 22:46:19
23/Mar/2007 14:40:07
23/Mar/2007 14:40:23
23/Mar/2007 15:58:46

Hi, it looks like your xrange argument needs to be enclosed in quotes on
the command line.  Also the format of your dates in the range argument
(dd-mmm-yyyy) is not the same as that in your data (dd/mmm/yyyy).

Steve


On Fri, 23 Mar 2007, g0sp3r wrote:

> Hello.
>
> I am trying to graph hits per hour using the chron prefab. I am
> running into trouble with specifying an xrange. My dataset spans
> severals days over several weeks. I want to graph all hits in a single
> day specified by my xrange= . Ploticus seems to try and graph every
> day present in my dataset instead of just the day I specified with
> xrange=. Anyone have any clues?
>
> I amm calling ploticus as follows
>
> pl -backcolor black  -color white  -png -o @ARGV[1] -prefab chron
> data=@ARGV[0] xrange= 22-mar-2007.0:0:0 22-mar-2007.23:59:59
> tab=hour  datefmt=dd/mmm/yyyy xmargin=0  xinc=\"1 hour\" barwidth=line
>  stubfmt=hh unittype=datetime timefld=2  title=\"\# hits per hour\"
> color=green
>
> my dataset is as follows:
>
> 19/Mar/2007 23:04:36
> 19/Mar/2007 23:05:27
> 19/Mar/2007 23:08:27
> 19/Mar/2007 23:08:40
> 19/Mar/2007 23:11:53
> 19/Mar/2007 23:11:58
> 19/Mar/2007 23:16:55
> 20/Mar/2007 01:10:10
> 20/Mar/2007 05:40:53
> 20/Mar/2007 05:43:57
> 20/Mar/2007 12:40:25
> 20/Mar/2007 19:22:04
> 20/Mar/2007 21:14:49
> 21/Mar/2007 01:26:09
> 21/Mar/2007 09:29:03
> 21/Mar/2007 14:26:34
> 21/Mar/2007 18:10:08
> 22/Mar/2007 03:01:19
> 22/Mar/2007 11:13:51
> 22/Mar/2007 11:19:22
> 22/Mar/2007 14:12:15
> 22/Mar/2007 19:45:46
> 22/Mar/2007 22:46:19
> 23/Mar/2007 14:40:07
> 23/Mar/2007 14:40:23
> 23/Mar/2007 15:58:46
>
>
>
>
>
>


   Stephen C. Grubb     scg@...                x-6633
   Scientific Software Engineer, The Jackson Laboratory
   600 Main Street  Bar Harbor, Maine 04609 USA

Hi.

I found some mistake in result image after running ploticus script.
Result image have big space under the plot.
I open "mydump" file and found points with negative Y coords.
May be that thing is a source of problem?
If I filter mydump without negative points, result image is ok.

Can you check this?

simple script:

#!/bin/sh

name=$0

pl -prefab vbars data=- yrange="35 55" x=1 y=2 -drawdump mydump \
     << EOF
1991 51.7
2050 36.6
EOF

#  cut negative numbers
#sed /-/d < mydump > dump
#  rename
#mv -f dump mydump

pl -prefab draw dumpfile=mydump -eps -o ${name}.eps


P.S. Please forgive me for my bad english :)

Hi, I was able to reproduce this problem here.

I believe you are right.. not sure why those Y<0 points are in there.

Thanks for finding & reporting this bug.  I will look into it shortly.
-Steve


---------- Forwarded message ----------
From: Koval Valery <Valery.Koval@...>
Date: Apr 11, 2007 3:07 AM
Subject: ploticus v2.33
To: stevegrubb@...

Hi.

I found some mistake in result image after running ploticus script.
Result image have big space under the plot.
I open "mydump" file and found points with negative Y coords.
May be that thing is a source of problem?
If I filter mydump without negative points, result image is ok.

Can you check this?

simple script:

#!/bin/sh

name=$0

pl -prefab vbars data=- yrange="35 55" x=1 y=2 -drawdump mydump \
      << EOF
1991 51.7
2050 36.6
EOF

#  cut negative numbers
#sed /-/d < mydump > dump
#  rename
#mv -f dump mydump

pl -prefab draw dumpfile=mydump -eps -o ${name}.eps


P.S. Please forgive me for my bad english :)

Michael, I was able to reproduce this problem here... thanks for finding
& reporting this bug.  I will look into it shortly.  -Steve

On Sat, 31 Mar 2007, Michael Butscher wrote:

> Hi,
>
> because I couldn't find a bug tracker on the Sourceforge page, I write
> the following directly to you.
>
> When using floating segment bars, it seems that Ploticus draws in a bar
> a thin vertical (for horizontal bars) line at the position where the
> previous bar (above the current) ended. I don't know how to remove this.
>
> You can see this for your example "timeline2.htm" ("Gallery of ploticus
> script examples" -> "Horizontal bars" -> "Example 8").
>
> In the documentation it is right, but the attached GIF file shows how it
> looks like when I render it.
>
> I'm using version 2.33 Windows binary on Windows 2000.
>
>
>
> Kind regards,
>
> Michael
>
>


   Stephen C. Grubb     scg@...                x-6633
   Scientific Software Engineer, The Jackson Laboratory
   600 Main Street  Bar Harbor, Maine 04609 USA

Mike, add the 'locfield' statement eg.

   #proc bars
    color: gray(0.8)
    barwidth: 0.3
    outline: no
    horizontalbars: yes
    segmentfields: c d
locfield: 1
   #endproc

this assigns bars to positions based on your first data field.

without this, it defaults to dropping bars into the first 3 positions
including 0.

-Steve


On Mon, 5 Mar 2007, Mike Karisok wrote:

> Hi
>   I want to draw horizontal bars with several milestones on it.
> Actually everything works fine, but I always have a bar at the bottom
> of the x axis.
> I tried since hours to get rid of it but with no success.
> Any help is very much appreciated.
>
>   Kind regards,
> Mike
>
>   #proc settings
>  units: cm
> #endproc
>   #proc getdata
>  fieldnameheader: yes
>  data:
>  a b       c      d      e       f
>  1 "item"    01/03/2007 31/12/2008 01/01/0001  ""
>  2 "item"    01/03/2007 31/12/2008 01/05/2007  "M1"
>  2 "item"    01/03/2007 31/12/2008 01/08/2007  "M2"
>  3 "item"     01/03/2007 31/12/2008 01/06/2008  "M3"
>  3 "item"      01/03/2007 31/12/2008 01/04/2007  "Planned1"
>  3 "item"      01/03/2007 31/12/2008 01/07/2007  "Planned2"
>  4 "item"    01/03/2007 31/12/2008 01/01/0001  ""
> #endproc
>   #proc areadef
>    title: my Title
>    rectangle: 3 2 12 12
>    xscaletype: date dd/mm/yyyy
>    xrange: 01/03/2007 31/12/2008
>    yautorange: categories
>    ycategories:
>    1
>    2
>    3
> #endproc
>   #proc bars
>    color: gray(0.8)
>    barwidth: 0.3
>    outline: no
>    horizontalbars: yes
>    segmentfields: c d
> #endproc
>   #proc yaxis
>    stubs: categories
>    labeldetails: adjust=-0.4,0
> #endproc
>   #proc scatterplot
>    xfield: e
>    yfield: a
>    labelfield: f offset=0.2
>    textdetails: adjust=0,0.3
>    symbol: shape=triangle style=filled fillcolor=grey(0.01) radius=0.15
> #endproc
>   // Must be invoked again in order to draw the legend
> #proc scatterplot
>   xfield: e
>   yfield: a
>   symbol: shape=triangle style=filled fillcolor=grey(0.01) radius=0.15
>   legendlabel: Milestone
> #endproc
>   #proc legend
>    location: 4 1
>    format: accross
> #endproc
>
>
>
>
> ---------------------------------
> Don't be flakey. Get Yahoo! Mail for Mobile and
> always stay connected to friends.


   Stephen C. Grubb     scg@...                x-6633
   Scientific Software Engineer, The Jackson Laboratory
   600 Main Street  Bar Harbor, Maine 04609 USA

Hi,

two recently reported minor bugs have been fixed.

1) floating segment bars sometimes had an extra line glitch (try the
gallery example timeline2.htm)

2) when bar graphs produced while using -drawdump, the dumpfile had some
wild draw locations causing later result to have large area of blank space
at bottom

Both of these bugs can be fixed by getting this file:

http://ploticus.sf.net/download/proc_bars.c

putting it into your ploticus/src directory, and doing:  make clean; make

These fixes will also be present in the next release.

-Steve

Hi, Steve. Here is a minor issue with titles. The workaround is easy,
though perhaps not obvious.

The following command:

pl -gif -prefab stack data=revs title="this is a test title for this
graph" \
	 "titledet= align=right" -x=1 -y=2

(where 'rev' is a data file with two columns of numeric data)

creates a graph with the title string aligned to the left side of the
graph,
rather than the right. Using 'align=left' causes the title to be
aligned to the
right.

I am using Ploticus 2.32 on Mac OS X (10.4.9).


Josh Simons

Thanks Josh.

About as bass-ackwards as it can be isn't it...

I have fixed this.  Take the following code, put it into a file called
"chunk_title", and then put that file into your ploticus/prefabs
directory, replacing the "chunk_title" file that is already there.
No need to rebuild anything.

-Steve

//--------------------------------------------
//// chunk included by prefabs to do title...
#if @title != ""
   #proc annotate
   #set XCENTER = $arith(@AREARIGHT-@AREALEFT/2.0)
   #set XCENTER = $arith(@AREALEFT+@XCENTER)
   #set TOP = $arith(@AREATOP+0.2)
   location: @XCENTER @TOP
   #ifspec titledet textdetails
   #if $def(titledet) = 1
     #if @titledet like "*align=left"
       location: @AREALEFT @TOP
     #elseif @titledet like "*align=right"
       location: @AREARIGHT @TOP
     #endif
   #endif
   text: @title
   #endproc
#endif
//---------------------------------------------



On Wed, 18 Apr 2007, Josh Simons wrote:

> Hi, Steve. Here is a minor issue with titles. The workaround is easy,
> though perhaps not obvious.
>
> The following command:
>
> pl -gif -prefab stack data=revs title="this is a test title for this
> graph" \
>  "titledet= align=right" -x=1 -y=2
>
> (where 'rev' is a data file with two columns of numeric data)
>
> creates a graph with the title string aligned to the left side of the
> graph,
> rather than the right. Using 'align=left' causes the title to be
> aligned to the
> right.
>
> I am using Ploticus 2.32 on Mac OS X (10.4.9).
>
>
> Josh Simons
>
>
>


   Stephen C. Grubb     scg@...                x-6633
   Scientific Software Engineer, The Jackson Laboratory
   600 Main Street  Bar Harbor, Maine 04609 USA