Fred,
I have attached the EPS file and the script the generate it. The
graphical portion of the script is very simple, as you may see.
Also, I have used the tool 'eps2eps' to convert the EPS file from NCL to
another EPS file and compared both files. See below the differences in
the headers:
EPS file from NCL (boxplotGrupos.eps) :
%!PS-Adobe-2.0 EPSF-2.0
%%Creator: NCAR GKS
%%CreationDate: Sep 17 2006 15:49:43
%%BoundingBox: (atend)
%%DocumentFonts: (atend)
%%EndComments
%%BeginProlog
EPS file converted with 'eps2eps' (boxplotGrupos2.eps) :
%!PS-Adobe-3.0 EPSF-3.0
%%BoundingBox: 0 0 469 559
%%HiResBoundingBox: 0.000000 0.000000 468.499927 558.499927
%.........................................
%%Creator: GNU Ghostscript 707 (epswrite)
%%CreationDate: 2007/04/04 08:44:07
%%DocumentData: Clean7Bit
%%LanguageLevel: 2
%%EndComments
%%BeginProlog
I'm using NCAR Command Language Version 4.2.0.a034 and eps2png 2.5 in a
machine:
Linux beagle.dea.inpe.br 2.6.17-1.2142_FC4 #1 Tue Jul 11 22:41:14 EDT
2006 i686 athlon i386 GNU/Linux
Mateus
Fred Clare escreveu:
>
> On Apr 3, 2007, at 8:03 PM, Mateus da Silva Teixeira wrote:
>
> Mateus,
>
>> Fred Clare escreveu:
>>>
>>> On Apr 3, 2007, at 7:39 PM, Mateus da Silva Teixeira wrote:
>>>
>>>> Dear NCL users,
>>>>
>>>> The Mini Graphics Manual says that EPS workstation contains a
>>>> bounding box.
>>>> However, when I generate an EPS file and try to convert it to PNG,
>>>> this tool gives an error message:
>>>>
>>>> $ eps2png -png -output teste.png boxplotGrupos.eps
>>>> No bounding box in boxplotGrupos.eps
>>>> -g must be followed by <width>x<height>
>>>> Problem creating teste.png for boxplotGrupos.eps
>>>>
>>>> And any PNG file is created.
>>>> Is the Graphics Manual wrong? Or am I doing something wrong?
>>>
>>> You should be getting a bounding box. Are you using
>>> gsn_open_wks with a first argument of "eps?"
>>
>> Yes, I'm using 'eps' as the first argument of gsn_open_wks!
>
> I just ran an example that uses "eps" and indeed the output
> PostScript contained a bounding box. Could you please send
> me your output file? Also, what version of NCL are you using?
> If your source is fairly simple, it might be helpful for me to have
> that too.
>
> Fred Clare
>
>
>>> Currently the recommended way to do this is to create a
>>> regular .ps file, then use "ps2epsi" to generate the eps
>>> file, and then use "convert" on that.
>>
>> Ok! I also use 'convert' to directly convert the EPS file to PNG.
>> But I wanted to know why eps2png gives this error, since the manual
>> says that there is a bounding box!
>>
>>>
>>> We have an *experimental* png driver that may work for you
>>> (it uses the X11 rasterizer). To give it a try, use "png" as the
>>> first argument to gsn_open_wks. This driver is not currently
>>> supported, so use are your own risk.
>>
>> It would be good having such a driver in the future.
>> I used 'png' as argument for gsn_open_wks now, but the result is bad.
>> The quality of the graphic is poor!
>>
>> Thanks
>> Mateus
>>
>>
>>>
>>> Fred Clare
>>>
>>>
>>>>
>>>> Thanks
>>>> Mateus
>>>>
>>>> _______________________________________________
>>>> ncl-talk mailing list
>>>> ncl-talk_at_ucar.edu
>>>> http://mailman.ucar.edu/mailman/listinfo/ncl-talk
>>>
>>>
>>
>> _______________________________________________
>> ncl-talk mailing list
>> ncl-talk_at_ucar.edu
>> http://mailman.ucar.edu/mailman/listinfo/ncl-talk
>
>
load "$NCARG_ROOT/lib/ncarg/nclscripts/csm/gsn_code.ncl"
load "$NCARG_ROOT/lib/ncarg/nclscripts/csm/gsn_csm.ncl"
load "$NCARG_ROOT/lib/ncarg/nclscripts/csm/contributed.ncl"
load "$NCARG_ROOT/lib/ncarg/nclscripts/csm/shea_util.ncl"
load "$NCARG_ROOT/lib/ncarg/nclscripts/myscripts/MyFuncs.ncl"
begin
; parâmetros p/ análise
anoi = 1960
anof = 2004
nmes = 12
ndia = 31
deltaAno = anof-anoi+1
; arquivos de entrada
; estes arquivos devem conter as estações (linhas)
; na mesma ordem.
repo = "sul/"
arqGrupos = "membros_Sul.txt"
arqDados = "ncl4cluster_Sul.dat"
arqEstInfo = "estacoesProntas_Sul.txt"
; informações dos arquivos
dadosDias = numAsciiCol(repo+arqDados)
dadosEsta = numAsciiRow(repo+arqDados)
nEstacoes = numAsciiRow(repo+arqEstInfo)
nEstCols = numAsciiCol(repo+arqEstInfo)
; coordenada de tempo
; as datas da coordenada indica o número de dias desde
; 01 Jan 1960, às 00. O atributo de unidade é especialmente
; útil para a obtenção da data em formato mais
; 'amigável'.
time = fspan(0,dadosDias-1,dadosDias)
time_at_units = "days since "+sprinti("%4i",anoi)+"-01-01 00:00:00"
; lendo dados
grupos = asciiread(repo+arqGrupos,-1,"integer")
estacoes = asciiread(repo+arqEstInfo,(/nEstacoes,nEstCols/),"float")
dados = asciiread(repo+arqDados,(/dadosEsta,dadosDias/),"float")
dados@_FillValue = -999.9
; total de grupos
maxGrupos = max(grupos)
print("Nro de grupos encontrados: "+maxGrupos)
seriesGrupos = new((/maxGrupos,dadosDias/),float,-999.9)
; loop que varre os grupos
do i=1,maxGrupos
; nro de membros do i-ésimo grupo e os respectivos dados
nroMembros = num(grupos.eq.i)
idxMembros = ind(grupos.eq.i)
dadosMembros = dados(idxMembros,:)
dadosMembros!0 = "estacoes"
dadosMembros!1 = "time"
dadosMembros&time = time
; média de todas as estações
seriesGrupos(i-1,:) = dim_avg_Wrap(dadosMembros(time|:,estacoes|:))
delete(nroMembros) ; necessário pois são recalculados a cada
delete(idxMembros) ; iteração do loop externo
delete(dadosMembros) ;
end do
seriesGrupos!0 = "grupos"
seriesGrupos!1 = "time"
seriesGrupos&grupos = ispan(1,maxGrupos,1)
seriesGrupos&time = time
; preparando informações para o boxplot
dataBoxPlot = new((/maxGrupos,5/),float)
dataBoxPlot(:,0) = dim_min(seriesGrupos)
dataBoxPlot(:,4) = dim_max(seriesGrupos)
do i=0,maxGrupos-1
dataBoxPlot(i,1) = quantile(seriesGrupos(i,:),0.25)
dataBoxPlot(i,2) = quantile(seriesGrupos(i,:),0.5)
dataBoxPlot(i,3) = quantile(seriesGrupos(i,:),0.75)
end do
posX = (/-50.,-25.,0.,25.,50./)
; abrindo workstation
wks = gsn_open_wks("eps",repo+"boxplotGrupos")
; recursos gráficos
analisa = True
analisa_at_tmXBLabels = (/"Grp 1","Grp 2"," Grp 3","Grp 4","Grp 5"/)
plot = boxplot(wks,posX,dataBoxPlot,False,analisa,False)
draw(plot) ; obrigatórios p/ bloxplot
frame(wks)
end
_______________________________________________
ncl-talk mailing list
ncl-talk_at_ucar.edu
http://mailman.ucar.edu/mailman/listinfo/ncl-talk
This archive was generated by hypermail 2.2.0 : Wed Apr 11 2007 - 08:36:47 MDT