pgfplots-features — Announcements of new releases

Re: [Pgfplots-features] Problem with reusing legend names?

[Christian F. <cfe...@go...>]

Hi Florian,

I am unaware of any problems as long as you do not externalize or 
reference the legends somewhere else. Note that the external library 
would probably run into problems in these cases.

Kind regards

Christian

Am 28.02.19 um 12:55 schrieb Florian Lindner:
> Hello,
>
> when I use a separated legend, I always do it like that:
>
> \begin{figure}
>    \centering
>    \begin{tikzpicture}
>      \begin{axis}[
>       legend to name=leg:foo]
> 	[...]
>      \end{axis}
>    \end{tikzpicture}
>    \ref{leg:foo}
> \end{figure}
>
> I never use \ref{leg:foo} outside of the same figure it was created. In this case, is it a problem to reuse the same name, e.g., leg:foo for every plot?
>
> Best Thanks,
> Florian
>
>
> _______________________________________________
> Pgfplots-features mailing list
> Pgf...@li...
> https://lists.sourceforge.net/lists/listinfo/pgfplots-features

-- 

Dr. Christian Feuersänger
Rheinbach

[Pgfplots-features] Problem with reusing legend names?

[Florian L. <mai...@xg...>]

Hello,

when I use a separated legend, I always do it like that:

\begin{figure}
  \centering
  \begin{tikzpicture}
    \begin{axis}[
     legend to name=leg:foo]
	[...]
    \end{axis}
  \end{tikzpicture}
  \ref{leg:foo} 
\end{figure}

I never use \ref{leg:foo} outside of the same figure it was created. In this case, is it a problem to reuse the same name, e.g., leg:foo for every plot?

Best Thanks,
Florian

Re: [Pgfplots-features] enlarge x limits in groupplot and multiple symbols in legend

[Stefan P. <Mo...@gm...>]

Hi Florian,

> thanks for your answer! Setting ymin did the trick on the first, area
> limits on my second question.
>
> So far, I have made good experiences with this mailing list, big
> thanks to eveybody who helped me! As you recommended, I tried
> Stackexchange for my next question, regarding pgfplotstable:
>
> https://tex.stackexchange.com/q/471089

Of course it is your decision where you ask the question ...
I have seen your question, but I doubt that this is (easily) possible which 
might be the reason that you didn't receive an answer yet.

If you don't have too much data I guess it would be easily possible to restate 
the data table as a `tabular` and use `siunitx` `S` column to achieve what you 
want. If you are really lucky you can use `pgfplotstable` to output stuff in 
an `S` column ...


Best regards,
Stefan

Re: [Pgfplots-features] enlarge x limits in groupplot and multiple symbols in legend

[Florian L. <mai...@xg...>]

Hi Stefan,

thanks for your answer! Setting ymin did the trick on the first, area limits on my second question.

So far, I have made good experiences with this mailing list, big thanks to eveybody who helped me! As you recommended, I tried Stackexchange for my next question, regarding pgfplotstable:

https://tex.stackexchange.com/questions/471089/mix-numbers-and-strings-in-same-column

Best Thanks,
Florian

Am 17.01.19 um 19:27 schrieb Stefan Pinnow:
> Hi Florian,
> 
> - Using PGFPlots v1.16 `enlarge x limits` is working fine on both
>   groupplots.
> - The simplest way of getting rid of the both bars as legend image is
>   to add `area legend` to the ` groupplot` options.
>   If you need to draw something else, have a look at section 4.9.5
>   page 260ff (especially page 262) of the current manual (v1.16)
> 	<http://mirrors.ctan.org/graphics/pgf/contrib/pgfplots/doc/pgfplots.pdf#page=262>
> 
> Besides that you maybe have noticed that the `ymin` values are different for the two plots. Thus I recommend setting `ymin=1` explicitly. This will then change also the first plot lower part because of your `enlarge y limits` statement. To make it look similar as before change the given value to something like 0.01.
> 
> Because you haven't stated a MWE, i.e. I don't know which `compat` level you are using: In case you use a value >1.5 the lower end of the bars will be cut at 10^0. To avoid this you should also add `log origin=infty` to the `groupplot` options.
> 
> 
> Best regards,
> Stefan
> 
> 
> PS: I recommend that you in future ask these kind of questions on
> 	<https://tex.stackexchange.com/>
> The chance of getting an answer there is much much greater than on this mailing list.
> 
> 
>> -----Original Message-----
>> From: Florian Lindner [mailto:mai...@xg...]
>> Sent: Wednesday, January 16, 2019 2:48 PM
>> To: pgf...@li...
>> Subject: [Pgfplots-features] enlarge x limits in groupplot and multiple
>> symbols in legend
>>
>> Hello,
>>
>> I have a plot, generated from the two tiny data files I have attached. It
>> basically works great, with two small issues:
>>
>> * enlarge x limits seems to have only an effect on the first group plot. I
>> also tried putting it in \nextgroupplot[here], but same effect.
>>
>> * The legend contains two symbols (of the same color) for each entry, like
>> "☐ ☐ Filling evaluation". How can I omit the second symbol?
>>
>> Thanks a lot!
>> Florian
>>
>> \begin{figure}
>>   \centering
>>   \ref{leg:TimingsPreallocation}
>>   \begin{tikzpicture}
>>     \begin{groupplot}[
>>       group style={columns=2, y descriptions at=edge left},
>>       ybar={0cm},
>>       symbolic x coords={off, compute, save, tree},
>>       xtick=data,
>>       xticklabels={No preallocation, Explicitly computed, Computed \&
>> saved, Spatial tree},
>>       /pgf/bar width=9pt,
>>       enlarge x limits=0.18, enlarge y limits={value=0.3, lower},
>>       ymode=log,
>>       ylabel={Time [ms]},
>>       x tick label style={rotate=30},
>>       ymax=40000,
>>       legend entries={Filling evaluation, Preallocation evaluation,
>> Filling interpolation, Preallocation interpolation},
>>       legend to name={leg:TimingsPreallocation}, legend columns=2]
>>
>>       \nextgroupplot
>>       \addplot+ table[y=fillA]{prealloc-80.csv};
>>       \addplot+ table[y=preallocA]{prealloc-80.csv};
>>       \addplot+ table[y=fillC]{prealloc-80.csv};
>>       \addplot+ table[y=preallocC]{prealloc-80.csv};
>>
>>       \nextgroupplot
>>       \addplot+ table[y=fillA]{prealloc-100.csv};
>>       \addplot+ table[y=preallocA]{prealloc-100.csv};
>>       \addplot+ table[y=fillC]{prealloc-100.csv};
>>       \addplot+ table[y=preallocC]{prealloc-100.csv};
>>     \end{groupplot}
>>   \end{tikzpicture}
>> \end{figure}

Re: [Pgfplots-features] enlarge x limits in groupplot and multiple symbols in legend

[Stefan P. <Mo...@gm...>]

Hi Florian,

- Using PGFPlots v1.16 `enlarge x limits` is working fine on both
  groupplots.
- The simplest way of getting rid of the both bars as legend image is
  to add `area legend` to the ` groupplot` options.
  If you need to draw something else, have a look at section 4.9.5
  page 260ff (especially page 262) of the current manual (v1.16)
	<http://mirrors.ctan.org/graphics/pgf/contrib/pgfplots/doc/pgfplots.pdf#page=262>

Besides that you maybe have noticed that the `ymin` values are different for the two plots. Thus I recommend setting `ymin=1` explicitly. This will then change also the first plot lower part because of your `enlarge y limits` statement. To make it look similar as before change the given value to something like 0.01.

Because you haven't stated a MWE, i.e. I don't know which `compat` level you are using: In case you use a value >1.5 the lower end of the bars will be cut at 10^0. To avoid this you should also add `log origin=infty` to the `groupplot` options.


Best regards,
Stefan


PS: I recommend that you in future ask these kind of questions on
	<https://tex.stackexchange.com/>
The chance of getting an answer there is much much greater than on this mailing list.


> -----Original Message-----
> From: Florian Lindner [mailto:mai...@xg...]
> Sent: Wednesday, January 16, 2019 2:48 PM
> To: pgf...@li...
> Subject: [Pgfplots-features] enlarge x limits in groupplot and multiple
> symbols in legend
> 
> Hello,
> 
> I have a plot, generated from the two tiny data files I have attached. It
> basically works great, with two small issues:
> 
> * enlarge x limits seems to have only an effect on the first group plot. I
> also tried putting it in \nextgroupplot[here], but same effect.
> 
> * The legend contains two symbols (of the same color) for each entry, like
> "☐ ☐ Filling evaluation". How can I omit the second symbol?
> 
> Thanks a lot!
> Florian
> 
> \begin{figure}
>   \centering
>   \ref{leg:TimingsPreallocation}
>   \begin{tikzpicture}
>     \begin{groupplot}[
>       group style={columns=2, y descriptions at=edge left},
>       ybar={0cm},
>       symbolic x coords={off, compute, save, tree},
>       xtick=data,
>       xticklabels={No preallocation, Explicitly computed, Computed \&
> saved, Spatial tree},
>       /pgf/bar width=9pt,
>       enlarge x limits=0.18, enlarge y limits={value=0.3, lower},
>       ymode=log,
>       ylabel={Time [ms]},
>       x tick label style={rotate=30},
>       ymax=40000,
>       legend entries={Filling evaluation, Preallocation evaluation,
> Filling interpolation, Preallocation interpolation},
>       legend to name={leg:TimingsPreallocation}, legend columns=2]
> 
>       \nextgroupplot
>       \addplot+ table[y=fillA]{prealloc-80.csv};
>       \addplot+ table[y=preallocA]{prealloc-80.csv};
>       \addplot+ table[y=fillC]{prealloc-80.csv};
>       \addplot+ table[y=preallocC]{prealloc-80.csv};
> 
>       \nextgroupplot
>       \addplot+ table[y=fillA]{prealloc-100.csv};
>       \addplot+ table[y=preallocA]{prealloc-100.csv};
>       \addplot+ table[y=fillC]{prealloc-100.csv};
>       \addplot+ table[y=preallocC]{prealloc-100.csv};
>     \end{groupplot}
>   \end{tikzpicture}
> \end{figure}

[Pgfplots-features] enlarge x limits in groupplot and multiple symbols in legend

[Florian L. <mai...@xg...>]

Hello,

I have a plot, generated from the two tiny data files I have attached. It basically works great, with two small issues:

* enlarge x limits seems to have only an effect on the first group plot. I also tried putting it in \nextgroupplot[here], but same effect.

* The legend contains two symbols (of the same color) for each entry, like "☐ ☐ Filling evaluation". How can I omit the second symbol?

Thanks a lot!
Florian

\begin{figure}
  \centering
  \ref{leg:TimingsPreallocation}
  \begin{tikzpicture}
    \begin{groupplot}[
      group style={columns=2, y descriptions at=edge left},
      ybar={0cm},
      symbolic x coords={off, compute, save, tree},
      xtick=data,
      xticklabels={No preallocation, Explicitly computed, Computed \& saved, Spatial tree},
      /pgf/bar width=9pt,
      enlarge x limits=0.18, enlarge y limits={value=0.3, lower}, 
      ymode=log,
      ylabel={Time [ms]},
      x tick label style={rotate=30},
      ymax=40000,
      legend entries={Filling evaluation, Preallocation evaluation, Filling interpolation, Preallocation interpolation},
      legend to name={leg:TimingsPreallocation}, legend columns=2]
      
      \nextgroupplot
      \addplot+ table[y=fillA]{prealloc-80.csv};
      \addplot+ table[y=preallocA]{prealloc-80.csv};
      \addplot+ table[y=fillC]{prealloc-80.csv};
      \addplot+ table[y=preallocC]{prealloc-80.csv};

      \nextgroupplot
      \addplot+ table[y=fillA]{prealloc-100.csv};
      \addplot+ table[y=preallocA]{prealloc-100.csv};
      \addplot+ table[y=fillC]{prealloc-100.csv};
      \addplot+ table[y=preallocC]{prealloc-100.csv};
    \end{groupplot}
  \end{tikzpicture}
\end{figure}

Re: [Pgfplots-features] Bug in pos argument to node?

[Stefan P. <Mo...@gm...>]

Hi Florian,

this is intended behavior. To prove that remove `mark=none' and add `only 
marks' to the `axis' options. An explanation can be found in the PGFPlots 
manual for the `/tikz/pos={<fraction>}' in section 4.17.2 on page 356.

And if you are thinking "snap-to-nearest" should also work for `\addplot's 
*with* lines, there already exists such a feature request ;)
	<https://sourceforge.net/p/pgfplots/feature-requests/61/>


Best regards,
Stefan


> -----Original Message-----
> From: Florian Lindner [mailto:mai...@xg...]
> Sent: Wednesday, October 31, 2018 3:25 PM
> To: pgf...@li...
> Subject: [Pgfplots-features] Bug in pos argument to node?
>
> Hey,
>
> I just recently discovered you can use addplot followed by a node[pos=0.3]
> for labeling plots. I have a tex file:
>
> \documentclass{scrartcl}
>
> \usepackage{tikz}
> \usepackage{pgfplots}
> \pgfplotsset{compat=1.16, table/col sep=comma}
>
> \begin{document}
>
> \begin{tikzpicture}[
>   label/.style={font=\small}]]
>   \begin{axis}[
>     grid=major, xmode=log, ymode=log,
>     x dir=reverse,
>     every axis plot post/.append style={x=h, mark=none}]
>
>     \addplot+ table[y=InfError]
> {h_convergence_NoneConsistent_MultiQuadrics_highfreq_14.csv} node[label,
> yshift=0.5cm, pos=0.3] {MQ 14};
>     \addplot+ table[y=InfError]
> {h_convergence_NoneConsistent_MultiQuadrics_highfreq_4.csv} node[label,
> yshift=-1cm, pos=0.3] {MQ 4};
>     \addplot+ table[y=InfError]
> {h_convergence_NoneConsistent_ThinPlateSplines_highfreq_0.csv} node[label,
> pos=0.3] {TPS};
>     \addplot+ table[y=InfError]
> {h_convergence_NoneConsistent_VolumeSplines_highfreq_0.csv} node[label,
> pos=0.3] {VS};
>
>   \end{axis}
> \end{tikzpicture}
>
> \end{document}
>
>
> The x column h is the same at all csv files, so I expect the nodes to be
> placed at the same x-position. But the attached PDF shows that the two MQ
> nodes are placed behind the VS/TPS nodes.
>
> It also happens without logarithmic axes.
>
> I have attached all data to reproduce the problem. Is that a bug in how
> pos is implemented or a misunderstanding on my side?
>
> Best Thanks,
> Florian

[Pgfplots-features] Bug in pos argument to node?

[Florian L. <mai...@xg...>]

Hey,

I just recently discovered you can use addplot followed by a node[pos=0.3] for labeling plots. I have a tex file:

\documentclass{scrartcl}

\usepackage{tikz}
\usepackage{pgfplots}
\pgfplotsset{compat=1.16, table/col sep=comma}

\begin{document}

\begin{tikzpicture}[
  label/.style={font=\small}]]
  \begin{axis}[
    grid=major, xmode=log, ymode=log,
    x dir=reverse,
    every axis plot post/.append style={x=h, mark=none}]

    \addplot+ table[y=InfError] {h_convergence_NoneConsistent_MultiQuadrics_highfreq_14.csv} node[label, yshift=0.5cm, pos=0.3] {MQ 14};
    \addplot+ table[y=InfError] {h_convergence_NoneConsistent_MultiQuadrics_highfreq_4.csv} node[label, yshift=-1cm, pos=0.3] {MQ 4};    
    \addplot+ table[y=InfError] {h_convergence_NoneConsistent_ThinPlateSplines_highfreq_0.csv} node[label, pos=0.3] {TPS};
    \addplot+ table[y=InfError] {h_convergence_NoneConsistent_VolumeSplines_highfreq_0.csv} node[label, pos=0.3] {VS};

  \end{axis}
\end{tikzpicture}

\end{document}


The x column h is the same at all csv files, so I expect the nodes to be placed at the same x-position. But the attached PDF shows that the two MQ nodes are placed behind the VS/TPS nodes.

It also happens without logarithmic axes.

I have attached all data to reproduce the problem. Is that a bug in how pos is implemented or a misunderstanding on my side?

Best Thanks,
Florian

Re: [Pgfplots-features] Filter data before plotting

[Florian L. <mai...@xg...>]

Hey Dilawar,

thanks for the hint with lualatex, it's actually significantly faster:

With filtering (dataset changed a bit from the one I used before)
- pdflatex: 80s
- lualatex: 20s

Without filtering (i.e. I split the CSV file before):
- pdflatex: 2s
- lualatex: 2s (generally a bit slower than pdflatex here)

I don't think it's the size of the dataset, but the filtering, especially because I apply four filters.

Another idea might be use use the `shell` functionality of pgfplots to have an external program filter the rows. Or just live with it and split up the csv before latex.

Best,
Florian


Am 30.10.18 um 11:44 schrieb Dilawar Singh:
> Hi Florian
> 
> I have used pgfplots with lualatex/gnuplot with much larger dataset. pdflatex
> runs out of memory pretty soon on large dataset. It is PROBABLY slower than
> lualatex. You can checkout the generated images and scripts to generate them
> here.
> 
>    https://github.com/dilawar/SinghAndBhalla_CaMKII_SubunitExchange_2018/tree/master/PaperFigures
> 
> Moreover, I use `gnuplot` with its 'every' keyword which is very helpful if you
> need not plot every datapoint. In the example below, I plot every 10th point. If
> you need to plot every point, then would  be of no help.
> 
>    \begin{tikzpicture}[scale=1]
>        \begin{axis}[ xlabel=xlabel,ylabel=ylabel ]
>            \addplot [color=blue] gnuplot [ raw gnuplot ] {
>                use datafile separator ",";
>                plot "data.csv" using 2:3 with lines every 10
>            };
>        \end{axis}
>    \end{tikzpicture}
> 
> The learning curve was bit steep for me but it paid well.
> cheers,
>    Dliawar
> 
> On Mon, Oct 29, 2018 at 02:26:12PM +0100, Florian Lindner wrote:
>> Hey,
>>
>> unfortunately I got to walk away from using pgfplots like that again. It's just to slow.
>>
>> I have a document with nothing but 2 groupplots with 2 graphs and 4 or 5 addplots each.
>>
>> All are plot from the same CSV file with 3600 lines, which is filtered by 3 onlyif commands to 50 data points per addplot.
>>
>> On my machine pdflatex on that documents take 95s. (CPU is i5 @ 1.9 GHz, SSD).
>>
>> So I will split up data in chunks of CSV files before plotting.
>>
>> Best,
>> Florian
>>
>> Am 25.10.18 um 10:13 schrieb Florian Lindner:
>>> Hi,
>>>
>>> ok, I see the issue.
>>>
>>> One other thing. I get tons of messages like
>>>
>>> NOTE: coordinate (0.6,0.1) has been dropped because it is unbounded (in x). (see also unbounded coords=jump).
>>>
>>> I added filter discard warning=false, so the log output is not cluttered so much:
>>>
>>> \pgfplotsset{
>>>   only if/.style 2 args={
>>>     filter discard warning=false,
>>>     x filter/.append code={
>>>       \edef\tempa{\thisrow{#1}}
>>>       \edef\tempb{#2}
>>>       \ifx\tempa\tempb
>>>       \else
>>>       \def\pgfmathresult{NaN}
>>>       \fi
>>>     },
>>>   },
>>> }
>>>
>>> (so this message was less of a question, more of documentation)
>>>
>>> Best Regards,
>>>
>>> Florian
>>>
>>>
>>>
>>> Am 24.10.18 um 20:29 schrieb Stefan Pinnow:
>>>> Hi Florian,
>>>>
>>>> I think there was already a discussion about an inclusion of such a feature. I
>>>> think the conclusion was that it is not a trivial task, because the elements
>>>> to filter could be of numerical or non-numerical type, and the data could come
>>>> from a table with and without headers. And many, many more things need to be
>>>> considered.
>>>>
>>>> Thus I don't have much hope that Christian -- the developer -- will find a
>>>> "global" solution to this problem in the near future. But to keep track of
>>>> that, consider adding a feature request to the PGFPlots Tracker
>>>>     <https://sourceforge.net/p/pgfplots/feature-requests/>
>>>> if there isn't already one in for that feature.
>>>>
>>>>
>>>> Best regards,
>>>> Stefan
>>>>
>>>>
>>>>> -----Original Message-----
>>>>> From: Florian Lindner [mailto:mai...@xg...]
>>>>> Sent: Wednesday, October 24, 2018 5:34 PM
>>>>> To: Stefan Pinnow
>>>>> Cc: pgf...@li...
>>>>> Subject: Re: [Pgfplots-features] Filter data before plotting
>>>>>
>>>>> Hi Stefan,
>>>>>
>>>>> thanks for that link, works like a charm. I renamed it to "only if" for
>>>>> brevity.
>>>>>
>>>>> Still, I think that this is a worthy feature for inclusion into pgfplot,
>>>>> don't?
>>>>>
>>>>> Best Thanks,
>>>>> Florian
>>>>>
>>>>> Am 23.10.18 um 17:36 schrieb Stefan Pinnow:
>>>>>> Hello Florian,
>>>>>>
>>>>>> this is already possible with a minor modification of the code.
>>>>>> Have a look at
>>>>>>     <https://tex.stackexchange.com/a/342529/95441>
>>>>>> I am pretty sure that this question was also already there for `discard
>>>>> if'
>>>>>> but after a quick search I couldn't find the answer.
>>>>>>
>>>>>> But I think you already know what needs to be done to combine/"append"
>>>>>> filters ;)
>>>>>>
>>>>>>
>>>>>> Best regards,
>>>>>> Stefan
>>>>>>
>>>>>>
>>>>>>> -----Original Message-----
>>>>>>> From: Florian Lindner [mailto:mai...@xg...]
>>>>>>> Sent: Tuesday, October 23, 2018 11:10 AM
>>>>>>> To: pgf...@li...
>>>>>>> Subject: [Pgfplots-features] Filter data before plotting
>>>>>>>
>>>>>>> Hello,
>>>>>>>
>>>>>>> it seems to be common demand to filter data that is to be plotted.
>>>>>>>
>>>>>>> As an example, you have a table:
>>>>>>>
>>>>>>> |   h | Model | Testset | Result |
>>>>>>> |-----+-------+---------+--------|
>>>>>>> |   1 | A     | X       |    0.1 |
>>>>>>> | 0.5 | A     | X       |    0.2 |
>>>>>>> | 0.2 | A     | X       |    0.3 |
>>>>>>> |   1 | B     | X       |   0.11 |
>>>>>>> | 0.5 | B     | X       |   0.12 |
>>>>>>> | 0.2 | B     | X       |   0.13 |
>>>>>>> |   1 | B     | Y       |    0.1 |
>>>>>>> | 0.5 | B     | Y       |   0.12 |
>>>>>>> | 0.2 | B     | Y       |    0.1 |
>>>>>>>
>>>>>>> Now you want to plot all results (x = h, y = Result) for Model=B,
>>>>>>> Testset=Y.
>>>>>>>
>>>>>>> There are some code snippets, like:
>>>>>>>
>>>>>>> https://tex.stackexchange.com/questions/66640/how-can-i-filter-select-
>>>>>>> data-float-text-from-a-table-and-plot-it
>>>>>>> https://tex.stackexchange.com/questions/98003/filter-rows-from-a-table
>>>>>>>
>>>>>>> but they only allow to filter for one criteria.
>>>>>>>
>>>>>>> As far as I see, pgfplots already have a number of
>>>>> filtering/restrictions
>>>>>>> possiblilities, but they all act on coordinates only.
>>>>>>>
>>>>>>> Or is there something like \addplot[restrict={Model==B and Testset=Y},
>>>>>>> x=h, y=Result] {my_data.csv}; ?
>>>>>>>
>>>>>>> What would be the official way of doing that?
>>>>>>>
>>>>>>> Best Thanks,
>>>>>>>
>>>>>>> Florian Lindner
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>> _______________________________________________
>>>>>>> Pgfplots-features mailing list
>>>>>>> Pgf...@li...
>>>>>>> https://lists.sourceforge.net/lists/listinfo/pgfplots-features
>>>>>>
>>>
>>>
>>> _______________________________________________
>>> Pgfplots-features mailing list
>>> Pgf...@li...
>>> https://lists.sourceforge.net/lists/listinfo/pgfplots-features
>>>
>>
>>
>> _______________________________________________
>> Pgfplots-features mailing list
>> Pgf...@li...
>> https://lists.sourceforge.net/lists/listinfo/pgfplots-features

Re: [Pgfplots-features] Filter data before plotting

[Dilawar S. <dil...@gm...>]

Hi Florian

I have used pgfplots with lualatex/gnuplot with much larger dataset. pdflatex
runs out of memory pretty soon on large dataset. It is PROBABLY slower than
lualatex. You can checkout the generated images and scripts to generate them
here.

    https://github.com/dilawar/SinghAndBhalla_CaMKII_SubunitExchange_2018/tree/master/PaperFigures

Moreover, I use `gnuplot` with its 'every' keyword which is very helpful if you
need not plot every datapoint. In the example below, I plot every 10th point. If
you need to plot every point, then would  be of no help.

    \begin{tikzpicture}[scale=1]
        \begin{axis}[ xlabel=xlabel,ylabel=ylabel ]
            \addplot [color=blue] gnuplot [ raw gnuplot ] {
                use datafile separator ",";
                plot "data.csv" using 2:3 with lines every 10
            };
        \end{axis}
    \end{tikzpicture}

The learning curve was bit steep for me but it paid well. 

cheers,
    Dliawar

On Mon, Oct 29, 2018 at 02:26:12PM +0100, Florian Lindner wrote:
>Hey,
>
>unfortunately I got to walk away from using pgfplots like that again. It's just to slow.
>
>I have a document with nothing but 2 groupplots with 2 graphs and 4 or 5 addplots each.
>
>All are plot from the same CSV file with 3600 lines, which is filtered by 3 onlyif commands to 50 data points per addplot.
>
>On my machine pdflatex on that documents take 95s. (CPU is i5 @ 1.9 GHz, SSD).
>
>So I will split up data in chunks of CSV files before plotting.
>
>Best,
>Florian
>
>Am 25.10.18 um 10:13 schrieb Florian Lindner:
>> Hi,
>>
>> ok, I see the issue.
>>
>> One other thing. I get tons of messages like
>>
>> NOTE: coordinate (0.6,0.1) has been dropped because it is unbounded (in x). (see also unbounded coords=jump).
>>
>> I added filter discard warning=false, so the log output is not cluttered so much:
>>
>> \pgfplotsset{
>>   only if/.style 2 args={
>>     filter discard warning=false,
>>     x filter/.append code={
>>       \edef\tempa{\thisrow{#1}}
>>       \edef\tempb{#2}
>>       \ifx\tempa\tempb
>>       \else
>>       \def\pgfmathresult{NaN}
>>       \fi
>>     },
>>   },
>> }
>>
>> (so this message was less of a question, more of documentation)
>>
>> Best Regards,
>>
>> Florian
>>
>>
>>
>> Am 24.10.18 um 20:29 schrieb Stefan Pinnow:
>>> Hi Florian,
>>>
>>> I think there was already a discussion about an inclusion of such a feature. I
>>> think the conclusion was that it is not a trivial task, because the elements
>>> to filter could be of numerical or non-numerical type, and the data could come
>>> from a table with and without headers. And many, many more things need to be
>>> considered.
>>>
>>> Thus I don't have much hope that Christian -- the developer -- will find a
>>> "global" solution to this problem in the near future. But to keep track of
>>> that, consider adding a feature request to the PGFPlots Tracker
>>> 	<https://sourceforge.net/p/pgfplots/feature-requests/>
>>> if there isn't already one in for that feature.
>>>
>>>
>>> Best regards,
>>> Stefan
>>>
>>>
>>>> -----Original Message-----
>>>> From: Florian Lindner [mailto:mai...@xg...]
>>>> Sent: Wednesday, October 24, 2018 5:34 PM
>>>> To: Stefan Pinnow
>>>> Cc: pgf...@li...
>>>> Subject: Re: [Pgfplots-features] Filter data before plotting
>>>>
>>>> Hi Stefan,
>>>>
>>>> thanks for that link, works like a charm. I renamed it to "only if" for
>>>> brevity.
>>>>
>>>> Still, I think that this is a worthy feature for inclusion into pgfplot,
>>>> don't?
>>>>
>>>> Best Thanks,
>>>> Florian
>>>>
>>>> Am 23.10.18 um 17:36 schrieb Stefan Pinnow:
>>>>> Hello Florian,
>>>>>
>>>>> this is already possible with a minor modification of the code.
>>>>> Have a look at
>>>>> 	<https://tex.stackexchange.com/a/342529/95441>
>>>>> I am pretty sure that this question was also already there for `discard
>>>> if'
>>>>> but after a quick search I couldn't find the answer.
>>>>>
>>>>> But I think you already know what needs to be done to combine/"append"
>>>>> filters ;)
>>>>>
>>>>>
>>>>> Best regards,
>>>>> Stefan
>>>>>
>>>>>
>>>>>> -----Original Message-----
>>>>>> From: Florian Lindner [mailto:mai...@xg...]
>>>>>> Sent: Tuesday, October 23, 2018 11:10 AM
>>>>>> To: pgf...@li...
>>>>>> Subject: [Pgfplots-features] Filter data before plotting
>>>>>>
>>>>>> Hello,
>>>>>>
>>>>>> it seems to be common demand to filter data that is to be plotted.
>>>>>>
>>>>>> As an example, you have a table:
>>>>>>
>>>>>> |   h | Model | Testset | Result |
>>>>>> |-----+-------+---------+--------|
>>>>>> |   1 | A     | X       |    0.1 |
>>>>>> | 0.5 | A     | X       |    0.2 |
>>>>>> | 0.2 | A     | X       |    0.3 |
>>>>>> |   1 | B     | X       |   0.11 |
>>>>>> | 0.5 | B     | X       |   0.12 |
>>>>>> | 0.2 | B     | X       |   0.13 |
>>>>>> |   1 | B     | Y       |    0.1 |
>>>>>> | 0.5 | B     | Y       |   0.12 |
>>>>>> | 0.2 | B     | Y       |    0.1 |
>>>>>>
>>>>>> Now you want to plot all results (x = h, y = Result) for Model=B,
>>>>>> Testset=Y.
>>>>>>
>>>>>> There are some code snippets, like:
>>>>>>
>>>>>> https://tex.stackexchange.com/questions/66640/how-can-i-filter-select-
>>>>>> data-float-text-from-a-table-and-plot-it
>>>>>> https://tex.stackexchange.com/questions/98003/filter-rows-from-a-table
>>>>>>
>>>>>> but they only allow to filter for one criteria.
>>>>>>
>>>>>> As far as I see, pgfplots already have a number of
>>>> filtering/restrictions
>>>>>> possiblilities, but they all act on coordinates only.
>>>>>>
>>>>>> Or is there something like \addplot[restrict={Model==B and Testset=Y},
>>>>>> x=h, y=Result] {my_data.csv}; ?
>>>>>>
>>>>>> What would be the official way of doing that?
>>>>>>
>>>>>> Best Thanks,
>>>>>>
>>>>>> Florian Lindner
>>>>>>
>>>>>>
>>>>>>
>>>>>> _______________________________________________
>>>>>> Pgfplots-features mailing list
>>>>>> Pgf...@li...
>>>>>> https://lists.sourceforge.net/lists/listinfo/pgfplots-features
>>>>>
>>
>>
>> _______________________________________________
>> Pgfplots-features mailing list
>> Pgf...@li...
>> https://lists.sourceforge.net/lists/listinfo/pgfplots-features
>>
>
>
>_______________________________________________
>Pgfplots-features mailing list
>Pgf...@li...
>https://lists.sourceforge.net/lists/listinfo/pgfplots-features

Re: [Pgfplots-features] Filter data before plotting

[Florian L. <mai...@xg...>]

Hey,

unfortunately I got to walk away from using pgfplots like that again. It's just to slow.

I have a document with nothing but 2 groupplots with 2 graphs and 4 or 5 addplots each.

All are plot from the same CSV file with 3600 lines, which is filtered by 3 onlyif commands to 50 data points per addplot.

On my machine pdflatex on that documents take 95s. (CPU is i5 @ 1.9 GHz, SSD).

So I will split up data in chunks of CSV files before plotting.

Best,
Florian

Am 25.10.18 um 10:13 schrieb Florian Lindner:
> Hi,
> 
> ok, I see the issue.
> 
> One other thing. I get tons of messages like
> 
> NOTE: coordinate (0.6,0.1) has been dropped because it is unbounded (in x). (see also unbounded coords=jump).
> 
> I added filter discard warning=false, so the log output is not cluttered so much:
> 
> \pgfplotsset{
>   only if/.style 2 args={
>     filter discard warning=false,
>     x filter/.append code={
>       \edef\tempa{\thisrow{#1}}
>       \edef\tempb{#2}
>       \ifx\tempa\tempb
>       \else
>       \def\pgfmathresult{NaN}
>       \fi
>     },
>   },
> }
> 
> (so this message was less of a question, more of documentation)
> 
> Best Regards,
> 
> Florian
> 
> 
> 
> Am 24.10.18 um 20:29 schrieb Stefan Pinnow:
>> Hi Florian,
>>
>> I think there was already a discussion about an inclusion of such a feature. I 
>> think the conclusion was that it is not a trivial task, because the elements 
>> to filter could be of numerical or non-numerical type, and the data could come 
>> from a table with and without headers. And many, many more things need to be 
>> considered.
>>
>> Thus I don't have much hope that Christian -- the developer -- will find a 
>> "global" solution to this problem in the near future. But to keep track of 
>> that, consider adding a feature request to the PGFPlots Tracker
>> 	<https://sourceforge.net/p/pgfplots/feature-requests/>
>> if there isn't already one in for that feature.
>>
>>
>> Best regards,
>> Stefan
>>
>>
>>> -----Original Message-----
>>> From: Florian Lindner [mailto:mai...@xg...]
>>> Sent: Wednesday, October 24, 2018 5:34 PM
>>> To: Stefan Pinnow
>>> Cc: pgf...@li...
>>> Subject: Re: [Pgfplots-features] Filter data before plotting
>>>
>>> Hi Stefan,
>>>
>>> thanks for that link, works like a charm. I renamed it to "only if" for
>>> brevity.
>>>
>>> Still, I think that this is a worthy feature for inclusion into pgfplot,
>>> don't?
>>>
>>> Best Thanks,
>>> Florian
>>>
>>> Am 23.10.18 um 17:36 schrieb Stefan Pinnow:
>>>> Hello Florian,
>>>>
>>>> this is already possible with a minor modification of the code.
>>>> Have a look at
>>>> 	<https://tex.stackexchange.com/a/342529/95441>
>>>> I am pretty sure that this question was also already there for `discard
>>> if'
>>>> but after a quick search I couldn't find the answer.
>>>>
>>>> But I think you already know what needs to be done to combine/"append"
>>>> filters ;)
>>>>
>>>>
>>>> Best regards,
>>>> Stefan
>>>>
>>>>
>>>>> -----Original Message-----
>>>>> From: Florian Lindner [mailto:mai...@xg...]
>>>>> Sent: Tuesday, October 23, 2018 11:10 AM
>>>>> To: pgf...@li...
>>>>> Subject: [Pgfplots-features] Filter data before plotting
>>>>>
>>>>> Hello,
>>>>>
>>>>> it seems to be common demand to filter data that is to be plotted.
>>>>>
>>>>> As an example, you have a table:
>>>>>
>>>>> |   h | Model | Testset | Result |
>>>>> |-----+-------+---------+--------|
>>>>> |   1 | A     | X       |    0.1 |
>>>>> | 0.5 | A     | X       |    0.2 |
>>>>> | 0.2 | A     | X       |    0.3 |
>>>>> |   1 | B     | X       |   0.11 |
>>>>> | 0.5 | B     | X       |   0.12 |
>>>>> | 0.2 | B     | X       |   0.13 |
>>>>> |   1 | B     | Y       |    0.1 |
>>>>> | 0.5 | B     | Y       |   0.12 |
>>>>> | 0.2 | B     | Y       |    0.1 |
>>>>>
>>>>> Now you want to plot all results (x = h, y = Result) for Model=B,
>>>>> Testset=Y.
>>>>>
>>>>> There are some code snippets, like:
>>>>>
>>>>> https://tex.stackexchange.com/questions/66640/how-can-i-filter-select-
>>>>> data-float-text-from-a-table-and-plot-it
>>>>> https://tex.stackexchange.com/questions/98003/filter-rows-from-a-table
>>>>>
>>>>> but they only allow to filter for one criteria.
>>>>>
>>>>> As far as I see, pgfplots already have a number of
>>> filtering/restrictions
>>>>> possiblilities, but they all act on coordinates only.
>>>>>
>>>>> Or is there something like \addplot[restrict={Model==B and Testset=Y},
>>>>> x=h, y=Result] {my_data.csv}; ?
>>>>>
>>>>> What would be the official way of doing that?
>>>>>
>>>>> Best Thanks,
>>>>>
>>>>> Florian Lindner
>>>>>
>>>>>
>>>>>
>>>>> _______________________________________________
>>>>> Pgfplots-features mailing list
>>>>> Pgf...@li...
>>>>> https://lists.sourceforge.net/lists/listinfo/pgfplots-features
>>>>
> 
> 
> _______________________________________________
> Pgfplots-features mailing list
> Pgf...@li...
> https://lists.sourceforge.net/lists/listinfo/pgfplots-features
> 

Re: [Pgfplots-features] Filter data before plotting

[Florian L. <mai...@xg...>]

Hi,

ok, I see the issue.

One other thing. I get tons of messages like

NOTE: coordinate (0.6,0.1) has been dropped because it is unbounded (in x). (see also unbounded coords=jump).

I added filter discard warning=false, so the log output is not cluttered so much:

\pgfplotsset{
  only if/.style 2 args={
    filter discard warning=false,
    x filter/.append code={
      \edef\tempa{\thisrow{#1}}
      \edef\tempb{#2}
      \ifx\tempa\tempb
      \else
      \def\pgfmathresult{NaN}
      \fi
    },
  },
}

(so this message was less of a question, more of documentation)

Best Regards,

Florian



Am 24.10.18 um 20:29 schrieb Stefan Pinnow:
> Hi Florian,
> 
> I think there was already a discussion about an inclusion of such a feature. I 
> think the conclusion was that it is not a trivial task, because the elements 
> to filter could be of numerical or non-numerical type, and the data could come 
> from a table with and without headers. And many, many more things need to be 
> considered.
> 
> Thus I don't have much hope that Christian -- the developer -- will find a 
> "global" solution to this problem in the near future. But to keep track of 
> that, consider adding a feature request to the PGFPlots Tracker
> 	<https://sourceforge.net/p/pgfplots/feature-requests/>
> if there isn't already one in for that feature.
> 
> 
> Best regards,
> Stefan
> 
> 
>> -----Original Message-----
>> From: Florian Lindner [mailto:mai...@xg...]
>> Sent: Wednesday, October 24, 2018 5:34 PM
>> To: Stefan Pinnow
>> Cc: pgf...@li...
>> Subject: Re: [Pgfplots-features] Filter data before plotting
>>
>> Hi Stefan,
>>
>> thanks for that link, works like a charm. I renamed it to "only if" for
>> brevity.
>>
>> Still, I think that this is a worthy feature for inclusion into pgfplot,
>> don't?
>>
>> Best Thanks,
>> Florian
>>
>> Am 23.10.18 um 17:36 schrieb Stefan Pinnow:
>>> Hello Florian,
>>>
>>> this is already possible with a minor modification of the code.
>>> Have a look at
>>> 	<https://tex.stackexchange.com/a/342529/95441>
>>> I am pretty sure that this question was also already there for `discard
>> if'
>>> but after a quick search I couldn't find the answer.
>>>
>>> But I think you already know what needs to be done to combine/"append"
>>> filters ;)
>>>
>>>
>>> Best regards,
>>> Stefan
>>>
>>>
>>>> -----Original Message-----
>>>> From: Florian Lindner [mailto:mai...@xg...]
>>>> Sent: Tuesday, October 23, 2018 11:10 AM
>>>> To: pgf...@li...
>>>> Subject: [Pgfplots-features] Filter data before plotting
>>>>
>>>> Hello,
>>>>
>>>> it seems to be common demand to filter data that is to be plotted.
>>>>
>>>> As an example, you have a table:
>>>>
>>>> |   h | Model | Testset | Result |
>>>> |-----+-------+---------+--------|
>>>> |   1 | A     | X       |    0.1 |
>>>> | 0.5 | A     | X       |    0.2 |
>>>> | 0.2 | A     | X       |    0.3 |
>>>> |   1 | B     | X       |   0.11 |
>>>> | 0.5 | B     | X       |   0.12 |
>>>> | 0.2 | B     | X       |   0.13 |
>>>> |   1 | B     | Y       |    0.1 |
>>>> | 0.5 | B     | Y       |   0.12 |
>>>> | 0.2 | B     | Y       |    0.1 |
>>>>
>>>> Now you want to plot all results (x = h, y = Result) for Model=B,
>>>> Testset=Y.
>>>>
>>>> There are some code snippets, like:
>>>>
>>>> https://tex.stackexchange.com/questions/66640/how-can-i-filter-select-
>>>> data-float-text-from-a-table-and-plot-it
>>>> https://tex.stackexchange.com/questions/98003/filter-rows-from-a-table
>>>>
>>>> but they only allow to filter for one criteria.
>>>>
>>>> As far as I see, pgfplots already have a number of
>> filtering/restrictions
>>>> possiblilities, but they all act on coordinates only.
>>>>
>>>> Or is there something like \addplot[restrict={Model==B and Testset=Y},
>>>> x=h, y=Result] {my_data.csv}; ?
>>>>
>>>> What would be the official way of doing that?
>>>>
>>>> Best Thanks,
>>>>
>>>> Florian Lindner
>>>>
>>>>
>>>>
>>>> _______________________________________________
>>>> Pgfplots-features mailing list
>>>> Pgf...@li...
>>>> https://lists.sourceforge.net/lists/listinfo/pgfplots-features
>>>

Re: [Pgfplots-features] Filter data before plotting

[Stefan P. <Mo...@gm...>]

Hi Florian,

I think there was already a discussion about an inclusion of such a feature. I 
think the conclusion was that it is not a trivial task, because the elements 
to filter could be of numerical or non-numerical type, and the data could come 
from a table with and without headers. And many, many more things need to be 
considered.

Thus I don't have much hope that Christian -- the developer -- will find a 
"global" solution to this problem in the near future. But to keep track of 
that, consider adding a feature request to the PGFPlots Tracker
	<https://sourceforge.net/p/pgfplots/feature-requests/>
if there isn't already one in for that feature.


Best regards,
Stefan


> -----Original Message-----
> From: Florian Lindner [mailto:mai...@xg...]
> Sent: Wednesday, October 24, 2018 5:34 PM
> To: Stefan Pinnow
> Cc: pgf...@li...
> Subject: Re: [Pgfplots-features] Filter data before plotting
>
> Hi Stefan,
>
> thanks for that link, works like a charm. I renamed it to "only if" for
> brevity.
>
> Still, I think that this is a worthy feature for inclusion into pgfplot,
> don't?
>
> Best Thanks,
> Florian
>
> Am 23.10.18 um 17:36 schrieb Stefan Pinnow:
> > Hello Florian,
> >
> > this is already possible with a minor modification of the code.
> > Have a look at
> > 	<https://tex.stackexchange.com/a/342529/95441>
> > I am pretty sure that this question was also already there for `discard
> if'
> > but after a quick search I couldn't find the answer.
> >
> > But I think you already know what needs to be done to combine/"append"
> > filters ;)
> >
> >
> > Best regards,
> > Stefan
> >
> >
> >> -----Original Message-----
> >> From: Florian Lindner [mailto:mai...@xg...]
> >> Sent: Tuesday, October 23, 2018 11:10 AM
> >> To: pgf...@li...
> >> Subject: [Pgfplots-features] Filter data before plotting
> >>
> >> Hello,
> >>
> >> it seems to be common demand to filter data that is to be plotted.
> >>
> >> As an example, you have a table:
> >>
> >> |   h | Model | Testset | Result |
> >> |-----+-------+---------+--------|
> >> |   1 | A     | X       |    0.1 |
> >> | 0.5 | A     | X       |    0.2 |
> >> | 0.2 | A     | X       |    0.3 |
> >> |   1 | B     | X       |   0.11 |
> >> | 0.5 | B     | X       |   0.12 |
> >> | 0.2 | B     | X       |   0.13 |
> >> |   1 | B     | Y       |    0.1 |
> >> | 0.5 | B     | Y       |   0.12 |
> >> | 0.2 | B     | Y       |    0.1 |
> >>
> >> Now you want to plot all results (x = h, y = Result) for Model=B,
> >> Testset=Y.
> >>
> >> There are some code snippets, like:
> >>
> >> https://tex.stackexchange.com/questions/66640/how-can-i-filter-select-
> >> data-float-text-from-a-table-and-plot-it
> >> https://tex.stackexchange.com/questions/98003/filter-rows-from-a-table
> >>
> >> but they only allow to filter for one criteria.
> >>
> >> As far as I see, pgfplots already have a number of
> filtering/restrictions
> >> possiblilities, but they all act on coordinates only.
> >>
> >> Or is there something like \addplot[restrict={Model==B and Testset=Y},
> >> x=h, y=Result] {my_data.csv}; ?
> >>
> >> What would be the official way of doing that?
> >>
> >> Best Thanks,
> >>
> >> Florian Lindner
> >>
> >>
> >>
> >> _______________________________________________
> >> Pgfplots-features mailing list
> >> Pgf...@li...
> >> https://lists.sourceforge.net/lists/listinfo/pgfplots-features
> >

Re: [Pgfplots-features] Filter data before plotting

[Florian L. <mai...@xg...>]

Hi Stefan,

thanks for that link, works like a charm. I renamed it to "only if" for brevity.

Still, I think that this is a worthy feature for inclusion into pgfplot, don't?

Best Thanks,
Florian

Am 23.10.18 um 17:36 schrieb Stefan Pinnow:
> Hello Florian,
> 
> this is already possible with a minor modification of the code.
> Have a look at
> 	<https://tex.stackexchange.com/a/342529/95441>
> I am pretty sure that this question was also already there for `discard if'
> but after a quick search I couldn't find the answer.
> 
> But I think you already know what needs to be done to combine/"append"
> filters ;)
> 
> 
> Best regards,
> Stefan
> 
> 
>> -----Original Message-----
>> From: Florian Lindner [mailto:mai...@xg...]
>> Sent: Tuesday, October 23, 2018 11:10 AM
>> To: pgf...@li...
>> Subject: [Pgfplots-features] Filter data before plotting
>>
>> Hello,
>>
>> it seems to be common demand to filter data that is to be plotted.
>>
>> As an example, you have a table:
>>
>> |   h | Model | Testset | Result |
>> |-----+-------+---------+--------|
>> |   1 | A     | X       |    0.1 |
>> | 0.5 | A     | X       |    0.2 |
>> | 0.2 | A     | X       |    0.3 |
>> |   1 | B     | X       |   0.11 |
>> | 0.5 | B     | X       |   0.12 |
>> | 0.2 | B     | X       |   0.13 |
>> |   1 | B     | Y       |    0.1 |
>> | 0.5 | B     | Y       |   0.12 |
>> | 0.2 | B     | Y       |    0.1 |
>>
>> Now you want to plot all results (x = h, y = Result) for Model=B,
>> Testset=Y.
>>
>> There are some code snippets, like:
>>
>> https://tex.stackexchange.com/questions/66640/how-can-i-filter-select-
>> data-float-text-from-a-table-and-plot-it
>> https://tex.stackexchange.com/questions/98003/filter-rows-from-a-table
>>
>> but they only allow to filter for one criteria.
>>
>> As far as I see, pgfplots already have a number of filtering/restrictions
>> possiblilities, but they all act on coordinates only.
>>
>> Or is there something like \addplot[restrict={Model==B and Testset=Y},
>> x=h, y=Result] {my_data.csv}; ?
>>
>> What would be the official way of doing that?
>>
>> Best Thanks,
>>
>> Florian Lindner
>>
>>
>>
>> _______________________________________________
>> Pgfplots-features mailing list
>> Pgf...@li...
>> https://lists.sourceforge.net/lists/listinfo/pgfplots-features
> 

Re: [Pgfplots-features] Filter data before plotting

[Stefan P. <Mo...@gm...>]

Hello Florian,

this is already possible with a minor modification of the code.
Have a look at
	<https://tex.stackexchange.com/a/342529/95441>
I am pretty sure that this question was also already there for `discard if'
but after a quick search I couldn't find the answer.

But I think you already know what needs to be done to combine/"append"
filters ;)


Best regards,
Stefan


> -----Original Message-----
> From: Florian Lindner [mailto:mai...@xg...]
> Sent: Tuesday, October 23, 2018 11:10 AM
> To: pgf...@li...
> Subject: [Pgfplots-features] Filter data before plotting
> 
> Hello,
> 
> it seems to be common demand to filter data that is to be plotted.
> 
> As an example, you have a table:
> 
> |   h | Model | Testset | Result |
> |-----+-------+---------+--------|
> |   1 | A     | X       |    0.1 |
> | 0.5 | A     | X       |    0.2 |
> | 0.2 | A     | X       |    0.3 |
> |   1 | B     | X       |   0.11 |
> | 0.5 | B     | X       |   0.12 |
> | 0.2 | B     | X       |   0.13 |
> |   1 | B     | Y       |    0.1 |
> | 0.5 | B     | Y       |   0.12 |
> | 0.2 | B     | Y       |    0.1 |
> 
> Now you want to plot all results (x = h, y = Result) for Model=B,
> Testset=Y.
> 
> There are some code snippets, like:
> 
> https://tex.stackexchange.com/questions/66640/how-can-i-filter-select-
> data-float-text-from-a-table-and-plot-it
> https://tex.stackexchange.com/questions/98003/filter-rows-from-a-table
> 
> but they only allow to filter for one criteria.
> 
> As far as I see, pgfplots already have a number of filtering/restrictions
> possiblilities, but they all act on coordinates only.
> 
> Or is there something like \addplot[restrict={Model==B and Testset=Y},
> x=h, y=Result] {my_data.csv}; ?
> 
> What would be the official way of doing that?
> 
> Best Thanks,
> 
> Florian Lindner
> 
> 
> 
> _______________________________________________
> Pgfplots-features mailing list
> Pgf...@li...
> https://lists.sourceforge.net/lists/listinfo/pgfplots-features

[Pgfplots-features] Filter data before plotting

[Florian L. <mai...@xg...>]

Hello,

it seems to be common demand to filter data that is to be plotted.

As an example, you have a table:

|   h | Model | Testset | Result |
|-----+-------+---------+--------|
|   1 | A     | X       |    0.1 |
| 0.5 | A     | X       |    0.2 |
| 0.2 | A     | X       |    0.3 |
|   1 | B     | X       |   0.11 |
| 0.5 | B     | X       |   0.12 |
| 0.2 | B     | X       |   0.13 |
|   1 | B     | Y       |    0.1 |
| 0.5 | B     | Y       |   0.12 |
| 0.2 | B     | Y       |    0.1 |

Now you want to plot all results (x = h, y = Result) for Model=B, Testset=Y.

There are some code snippets, like:

https://tex.stackexchange.com/questions/66640/how-can-i-filter-select-data-float-text-from-a-table-and-plot-it
https://tex.stackexchange.com/questions/98003/filter-rows-from-a-table

but they only allow to filter for one criteria.

As far as I see, pgfplots already have a number of filtering/restrictions possiblilities, but they all act on coordinates only.

Or is there something like \addplot[restrict={Model==B and Testset=Y}, x=h, y=Result] {my_data.csv}; ?

What would be the official way of doing that?

Best Thanks,

Florian Lindner

[Pgfplots-features] pgfplots 1.16 released

[Christian F. <cfe...@go...>]

Hello users of pgfplots,

I am pleased to announce the release of pgfplots 1.16 featuring 
bugfixes, primarily intended to improve the stability and reliability of 
the 'fill between' library for unsmooth plots.

Kind regards

Christian

HISTORY:
1.16:
- fixed bug #111 [fillbetween] strange behavior when `soft clip` is used 
and one of the paths touches the axis border
- fixed bug #183: Nan in the first line of a numeric table was 
interpreted as column name
- fixed bug #187 Wrong output from mod in axis
- fixed bug #109 `visualization depends on` doesn't work with table from 
pgfplotstable is now fixed for numeric values
- fixed bug: \pgfplotstablenew was unable to create tables with 0 or 1 rows.
- fixed bug: 114 fillbetween suffered from inaccuracy (produced 
invisible segments)
- fixed bug #139 [fillbetween] numerical issues with dense points
- fixed bug #153: \begin{tikzpicture}[scale=...] combined with fill 
between or
   'set layers' resulted in a wrong bounding box
- \addplot gnuplot: autodetected unbounded coordinates
- new feature: parse tick positions using math parser (tracker #69)
- new feature: programmatic access to axis coordinates for given canvas 
coordinates (\pgfplotspointgetcoordinates, tracker 68)
- new feature: 'x filter/.append expression', a stackable variant of 'x 
filter/.expression'

1.15: bugfix release
- fixed regression: fpu caused forest library to fail as soon as one 
loads pgfplots
- fixed bug #149 : \edef{...\to...} was wrong and caused 
\pgfplotstablevertcat to fail
- usability: added support for \plotnum during \addplot
- fixed bug 134 [statistics] Histogram with custom `symbolic coords`
- fixed bug 140 (pgfplotstable read does not process empty rows anymore)
- fixed bug 150 Usage of \pgfkeysvalueof in xtick leads to 'Dimension 
too large'
- fixed bug 105 [manual] `xticklabel pos=upper` is not documented
     in fact, some of the documented positions for 'xtick pos' where
     unavailable.
- fixed bug 155 [groupplots] `scaled ticks` does not recognize 
`ticklabels at`
- fixed bug 160 `xticklabels` also used for `extra x tick labels` if 
they are not given explicitely
- implemented partial solution for bug 154 [log mode] sampling in 2D and 
3D different
- fixed bug 163 [bar-chart] bars dissapear when values are >xmax/ymax
- fixed bug 81 cannot use dollar sign as 'comment char'
- fixed bug 91 Hashes in data
- fixed bug 165 [minor ticks] minor tick drawn after the last xtick
- fixed bug 176: 2d plot expression in 3d axis fails for lualatex

1.14
- new feature: 'contour filled' (compat=1.14)
- new feature: building colormaps from other colormaps (see "Building 
Colormaps based on other Colormaps" in the manual)
- new feature: non-uniform colormaps (compat=1.14)
- new feature: colormaps defined on position of arbitrary magnitude
- new feature: colorbar as legend
- new feature: 'colorbar style={xtick=data}' positions tick labels at 
colormap positions
- fixed bug: pgfplots now handles incompatible changes of luatex
      loading \usepackage{pgfplots} _before_ pgf also allows makes older
      PGF versions compatible with luatex
- fixed bug: incompatibility between units + groupplots (bug 119)
- fixed bug: 'axis line shift' did not respect labels
- fixed bug: layers for axis lines were not respected
- fixed bug: two axes with fillbetween in the same picture failed due to 
clip paths on layers
- fixed bug: quiver plots with 'every arrow' failed to evaluate arrow 
tip length arguments
- fixed bug: \usepgfplotslibrary{colorbrewer}: colormap 'PuOr' was 
defined in reverse order

[Pgfplots-features] Release of pgfplots 1.15

[Christian F. <cfe...@go...>]

Hello users of pgfplots,

I am glad to announce the release of pgfplots 1.15 . It addressed a 
couple of issues found in the last months and is available for download 
on https://www.ctan.org/pkg/pgfplots and on 
https://sourceforge.net/projects/pgfplots . Package managers like tex 
live and miktex will be updated within the next days.

Kind regards

Christian

pgfplots - Create normal/logarithmic plots in two and three dimensions 
for LaTeX/TeX/ConTeXt.

pgfplotstable - Loads, rounds, formats and postprocesses numerical tables.

PGFPlots draws high--quality function plots in normal or logarithmic scaling
with a user-friendly interface directly in TeX. The user supplies axis 
labels,
legend entries and the plot coordinates for one or more plots and PGFPlots
applies axis scaling, computes any logarithms and axis ticks and draws the
plots. It supports line plots, scatter plots, piecewise constant plots, bar
plots, area plots, mesh and surface plots, patch plots, contour plots, 
quiver
plots, histogram plots, polar axes, ternary diagrams, smith charts and some
more.

Pgfplots is based on Till Tantau's package PGF/TikZ (pgf).

Pgfplotstable displays numerical tables rounded to desired precision in 
various
display formats, for example scientific format, fixed point format or 
integer,
using TeX's math facilities for pretty printing. Furthermore, it provides
methods for table postprocessing.

Please take a look at
   doc/latex/pgfplots/pgfplots.pdf
and
   doc/latex/pgfplots/pgfplotstable.pdf.

Copyright 2007-2017 by Christian Feuersaenger.

This program is free software: you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation, either version 3 of the License, or
(at your option) any later version.

This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
GNU General Public License for more details.

You should have received a copy of the GNU General Public License
along with this program.  If not, see <http://www.gnu.org/licenses/>.


HISTORY:
1.15: bugfix release
- fixed regression: fpu caused forest library to fail as soon as one 
loads pgfplots
- fixed bug #149 : \edef{...\to...} was wrong and caused 
\pgfplotstablevertcat to fail
- usability: added support for \plotnum during \addplot
- fixed bug 134 [statistics] Histogram with custom `symbolic coords`
- fixed bug 140 (pgfplotstable read does not process empty rows anymore)
- fixed bug 150 Usage of \pgfkeysvalueof in xtick leads to 'Dimension 
too large'
- fixed bug 105 [manual] `xticklabel pos=upper` is not documented
     in fact, some of the documented positions for 'xtick pos' where
     unavailable.
- fixed bug 155 [groupplots] `scaled ticks` does not recognize 
`ticklabels at`
- fixed bug 160 `xticklabels` also used for `extra x tick labels` if 
they are not given explicitely
- implemented partial solution for bug 154 [log mode] sampling in 2D and 
3D different
- fixed bug 163 [bar-chart] bars dissapear when values are >xmax/ymax
- fixed bug 81 cannot use dollar sign as 'comment char'
- fixed bug 91 Hashes in data
- fixed bug 165 [minor ticks] minor tick drawn after the last xtick
- fixed bug 176: 2d plot expression in 3d axis fails for lualatex

1.14
- new feature: 'contour filled' (compat=1.14)
- new feature: building colormaps from other colormaps (see "Building 
Colormaps based on other Colormaps" in the manual)
- new feature: non-uniform colormaps (compat=1.14)
- new feature: colormaps defined on position of arbitrary magnitude
- new feature: colorbar as legend
- new feature: 'colorbar style={xtick=data}' positions tick labels at 
colormap positions
- fixed bug: pgfplots now handles incompatible changes of luatex
      loading \usepackage{pgfplots} _before_ pgf also allows makes older
      PGF versions compatible with luatex
- fixed bug: incompatibility between units + groupplots (bug 119)
- fixed bug: 'axis line shift' did not respect labels
- fixed bug: layers for axis lines were not respected
- fixed bug: two axes with fillbetween in the same picture failed due to 
clip paths on layers
- fixed bug: quiver plots with 'every arrow' failed to evaluate arrow 
tip length arguments
- fixed bug: \usepgfplotslibrary{colorbrewer}: colormap 'PuOr' was 
defined in reverse order

1.13:
- fixed bug: incompatiblity between fillbetween and babel
- fixed bug: 'compat=1.9' (or newer) failed to work with log bar plots
- fixed bug: javascript incompatibility between Acrobat reader DC and 
clickable lib
- fixed bug in polar axes: repaired default axis label positions
- fixed bug in polar axes: repaired support for sloped descriptions
- fixed bug: 'axis line style={draw=none}' had no effect unless one had 
2d boxed axes
- fixed bug: point meta expressions with '=' inside of them which caused 
compilation errors
- fixed bug: compat=1.12 combined with interrupted plots (by empty line) 
failed to work
- fixed bug: compat=1.12 combined with lualatex evaluated relational 
operations (<, ==) with the wrong operator precedence
- fixed bug: fixed floor and ceil functions
- fixed bug: 'set layers' broke alignment features with different 
manifestations
        ('set layers' together with fill between in group plots, 
together with at={} key, together with anchor)
- fixed bug: fill between and group plots: fill-between graphics was 
missing (list of layers was lost)
- fixed bug: \addplot graphics {\newFile{P.pdf}}; failed to properly 
expand the file name
- fixed bug: fill between + error bars resulted in compilation error
- fixed bug: stacked plots occasionally used a wrong coordinate mapping
- fixed bug: provided suitable default legends for xbar and xbar stacked
- fixed bug: 'dateplot' now computes missing 'date ZERO' automatically. 
'date ZERO' is managed automatically.
- new feature: Added support for search paths to find data files 
(table/search path)
- new feature: 'xtick distance' allows to define tick positions by distance
- new feature: "axis line shift=10pt" shifts axis lines together with 
all their descriptions
- new feature: 'matrix plot' (alias 'imagesc') which allows cell-based 
matrix plots
- new feature: unbounded point meta in mesh plots will be treated like 
an invisible patch
- new feature: added high quality colormap 'viridis' (use 'colormap 
name=viridis' to activate it)
- new feature: ensured that 'shader=flat' returns the same result for 
every choice of 'z buffer' (starting with compat=1.13)
- new feature: utility function \pgfplotsforeachungrouped: added support 
for \breakforeach
- new feature: compat=1.13: activates sampling routine with less 
rounding errors
- new feature: cycle multiindex list (uses the 'ith' element for every 
provided sub-list)
- new feature: 'color of colormap=<value>' or 'color of colormap=<value> 
of <colormap name>'
- new feature: 'index of colormap=<value>' or 'index of colormap=<value> 
of <colormap name>'
- new feature: cycle list={[of colormap]}
- new feature: cycle list={[colors of colormap={0,100,500,100} of viridis]}
- new feature: cycle list={[indices of colormap={0,1,2,3} of hot]}
- new feature: added support for 'stack negative=on previous|separate' 
which allows separate plots for '+' and '-' (useful for bar plots)
- new feature: added support for 'colormap access=direct' combined with 
'shader=flat mean' and 'shader=interp':
     it previously used to interpolate indices instead of colors
- new feature: \closedcycle now produces useful results for 3d line plots
- new feature: added \usetikzlibrary{colorbrewer} and 
\usepgfplotslibrary{colorbrewer}
- new feature: added 'colormap access=piecewise constant'
- new feature: added rudimentary support for "Filled Contours" (based on 
'colormap access=piecewise constant')
- new feature: added methods to build colormaps based on other colormaps 
using
    "color of colormap", "index of colormap", "colors of colormap", 
"samples of colormap", and "indices of colormap"

1.12.1: minor bugfix release
- fixed bug: incompatibility of 1.12 with unicode-math
- fixed bug: lua backend failed to work with table input and dependency 
on \coordindex
- fixed bug: lua backend failed with 'forget plot'

1.12: scalability + performance
- scalability: reduced TeX memory consumption for huge inline tables
- scalability: lualatex allows bigger 3d graphics
- scalability: lualatex now reduces compilation time for huge 3d graphics
    (requires \pgfplotsset{compat=newest} at the time of this writing)
        surf shader=faceted: time reduced to 61% of pdflatex
         (\addplot3[surf,samples=70] {exp(-x^2-y^2 +x*y)})

     surf shader=interp: time reduced to 25% of pdflatex
          (\addplot3[surf,shader=interp,samples=70] {exp(-x^2-y^2 +x*y)})

     surf shader=interp, parametric: time reduced to 36% of pdflatex
         (\addplot3[surf,shader=interp,z buffer=sort, 
samples=50,domain=-1:0,y domain=0:2*pi] ({sqrt(1-x^2) * cos(deg(y))}, 
{sqrt( 1-x^2 ) * sin(deg(y))}, x);)

     surf shader=faceted, parametric: time reduced to 40% of pdflatex
         (\addplot3[surf,z buffer=sort, samples=50,domain=-1:0,y 
domain=0:2*pi] ({sqrt(1-x^2) * cos(deg(y))}, {sqrt( 1-x^2 ) * 
sin(deg(y))}, x);)

- scalability: lualated now reduces compilation time for boxplots.
     10000 input points from table: time reduced to 23% of pdflatex
- fixed bug: 3d axes where all axes used log scale produced compile failure
- fixed bug: axis equal did not respect "disabledatascaling"
- fixed bug: mesh legend was confused by color input=explicit
- fixed bug: colorbar style did not reset zmin/zmax
- fixed bug: memory allocation in FPU consumed too much memory
- fixed bug: boxplots: removed compilation error due to spurious spaces 
in options list
- fixed bug: boxplots: now uses standard quantile estimators: has all 9 
estimators of R and uses the same as in Excel as default (as of compat=1.12)
- fixed bug: boxplots: sometimes failed to compute boxplot statistics 
for special inputs
- fixed bug: boxplots: now, it sets 'mark=*' for outliers if there is no 
mark (compat=1.12)
- fixed regression: compat=1.11 broke arcs inside of axis
- new feature: ensured that arcs inside of axis make use of 'axis cs' 
and 'axis direction cs' automatically
- new feature: ensured that (0:1) is evaluated with axis cs
- fixed bug: precision of transformation cartesian to polar was too low 
(added atan2 to fpu)
- fixed bug: z buffer caused 'mesh legend' to produce a compile error
- new feature: added 'y filter/.expression={y==3 ? nan : y}' to simplify 
coordinate filters

1.11: usability + bug fixes.
- new feature: "hide obscured x ticks=false" to disable the feature 
which hides ticks
- simplified UI for symbolic coords: now, it accepts unexpandable
   dictionary entries including umlauts, undefined macros, math mode, ...
- simplified UI for symbolic coords: added magic prefix
     [normalized]<number> to allow something like xmin={[normalized]0.7}
- simplified UI for custom annotations: one does not need to say "axis 
cs" any longer;
   coordinates like (1,2) are now automatically interpreted as (axis cs:1,2)
- new feature: 'type ticklabels with struts' improves baseline alignment 
of tick labels
- new feature: 'trig format plots=rad' allows to use radians in 
trigonometric evaluations
- simplified 'external lib': added driver auto-detection for pdflatex, 
lualatex, xelatex, dvips
     (\usepgfplotslibrary{external}, or, with the next PGF release, 
\usetikzlibrary{external})
- simplified UI for \addplot coordinates: now, it expands macros as it goes.
- new feature: 'intersection segments' now accepts negative indices to 
count from the last backwards
- simplified UI: 'intersection segments' now accepts new syntax with 
1-based indices (old 0-based are still supported)
- fixed bug: fillbetween has had a problem with pins on input paths
- fixed bug: now, 'stacked ignores zero' is no longer applied to error bars.
- fixed bug: tick lines were drawn incorrectly for centered axis lines in 3d
- fixed bug: variable=u failed to work in some contexts
- fixed bug: fill between implicitly activated layers, but did not 
install the
    correct layers for grid lines etc.
- fixed bug: z grid style was applied to y grid style
- fixed bug: using 'x=1cm' failed to work in axis descriptions (legends)
- fixed bug: mesh legend failed to work with 'shader=interp'
- fixed bug: bounding box of shader=interp was wrong if there were 
transformations.
- fixed bug: added support for \addplot graphics in ConTeXt
- fixed bug: external lib failed to work with dvips (since PGF 3.0.0 I 
suppose); now it produces correct bounding boxes
- fixed bug: \pgfplotstablenew can now overwrite existing tables.
- fixed bug: extra ticks near intersections of axis lines were hidden
- fixed bug: 'intersection segments' did not work with '--cycle'
- fixed bug: 'intersection segments' did not update the last tikz point
- fixed bug: tick lines/grid lines have been clipped when they were near 
the boundary
- fixed bug: soft clip decoration failed to apply the transformation matrix
- fixed bug: polar axes and units library resulted in broken bounding box
- fixed bug: \addplot[stack plots=false] used to work up to 1.8, but 
failed until now. Fixed.
- fixed bug: hist/density was broken since some time


1.10:
- new feature: fill between plots (library fillbetween)
- new feature: concatenate intersection segments (library fillbetween)
- fixed bug: xelatex failed to run contour external
- fixed incompatibility with \label and \usepackage{mcaption}
- fixed bug: histograms produced wrong point meta
- fixed bug: histograms reported the wrong 'plot name' and confused 
shifts of bar plots

1.9:
- new feature: asymmetric error bars
- new feature: activated math parser for axis limit arguments, arguments 
in axis cs, and domain argument in log plots
- new feature: stacked bar plots place their 'nodes near coords' 
correctly in the middle and print the increment (compat=1.9)
- new feature: stacked bar plots suppress empty increments (compat=1.9).
- new feature: 'scatter/position=relative|absolute' allow to position 
'nodes near coords' absolutely.
    use-case: bar plots + nodes near coords which are at, say, y=0 
rather than their y value
- new feature: integration of smooth shadings & auto-CMYK conversion
     \usepackage[cmyk]{xcolor} or \selectcolormodel{cmyk} will 
reconfigure pgfplots to use CMYK (document-wide)
- new feature (advanced audience only): programmatic access to data 
coordinates during the visualization phase
     -> allows much more customization for error bars, stacked plots, 
nodes near coords.
- wrote beginner tutorials
- fixed bug: error bars and point meta did not work together
- fixed bug: stacked plots did not respect 'visualization depends on'
- fixed bug: luatex 0.76 is not backwards compatible; added version switch
- fixed bug: ternary library precision has been improved
- fixed bug: problem with axis limits very close to 0
- fixed bug: colormap specification limit case produced out of bounds 
exception


1.8:
- new feature: tight bounding box even if the axis is no box and bb 
excludes clip path
- new feature: mesh/color input=explicit
- new feature: shader=interp now has drivers for both dvipdfmx and xetex
- new feature: support for more color spaces in colormap definitions
- new feature: shader=interp and device-level gray colorspaces
- new feature: 'contour/contour dir=[xyz]' to draw contours in different 
directions
- new feature: statistics library with boxplot handler (both boxplot 
prepared and automatic computation)
- fixed bug: 3d centered axis lines and label placement (requires 
compat=1.8 or higher)
- fixed bug: axis lines and placement of labels, tick scale labels, and 
reversed axes
     (requires compat=1.8 or higher)
- fixed bug: filtering out coords from a mesh plot failed
- fixed bug: every legend image post was not respected inside of 
\ref{plotlabel}
- fixed bug: high-order patches computed the shader=flat mean in a wrong 
way.
- fixed bug: remember picture inside of pgfplots axes failed (due to 
cell picture)
- fixed bug: now, the tick scale label will be omitted if there are no ticks
- fixed bug: axis box path was not closed
- fixed bug: the bounding box was non-empty even if the axis was hidden.
- fixed bug: auto-alignment of nodes near coords failed for xbar plots
- fixed bug: providing bar width / bar shift in terms of axis units did
     not work with [xy]bar and nodes near coords
- fixed bug: transformation 'data cs=cart' -> polar is more robust now
- fixed bug: code did not compile against pgf 2.00
- fixed bug: patch plot lib and shader=interp,patch type=biquadratic
- fixed bug: context path searching issue (pgfplots.lua)
- fixed bug: shader=interp and dvips driver
- fixed bug: error bars with explicit relative input failed

1.7
- added feature: 'bar shift' and 'bar width' can now be expressed in 
terms of axis units (compat=1.7 or higher)
- fixed incompatibility regression pgfplots 1.6.1 pgf 2.10: layers
- fixed incompatibility pgfplots and imakeidx
- added feature: 'enlargelimits={abs=1cm}', i.e. enlarge by dimension 
rather than unit
- patchplots lib: added patch type=bicubic
- patchplots lib: added support for global paths (fillable)
- patchplots lib: added patch type sampling feature
- patchplots lib: improved usability (documentation and improvements)
- fixed path issues in context: moved lua input file to tex/generic
- fixed bug: \ref{legendimage} inside of legend text was wrong.

1.6.1:
- fixed incompatibility lualatex,shader=interp, and german package 
(introduced in 1.6)

1.6:
- added support for layered graphics (main use case: multiple axes and 
layers)
- added support for second colormap in mesh plots (mesh/interior 
colormap name)
- added support for scopes inside of axes
- contour plots: added ability to provide list of discrete labels 
(mesh/levels)
- empty lines are interpreted as interruptions in data plots (was 
undocumented since 1.4)
- added more scaling options to 'scale mode=scale uniformly' (affects 
axis equal in 3d
   and \addplot3 graphics)
- fixed wrong implementation of 'axis equal' and 'unit vector ratio' in 3d
   (backwards compatible for 2d, but not for 3d - the 3d implementation 
was plain wrong)
- fixed incompatibility of lualatex and shader=interp
- fixed bugs/added features around \addplot3 graphics
- fixed bug: colorbar did not support ymode=log
- fixed a couple of minor bugs
- fixed bounding box computation for clip=false,axis lines=none

1.5.1:
- more operations for FPU library (==, !=,<=,>=,?)
- fixed bug in usage of decorations in \addplot
- bugfix for contour prepared format=matlab
- added 'const plot mark mid' and 'jump mark mid' plot handlers
- nodes on a plot (\addplot ... node[pos=<fraction>] {};)
- 'trim axis group left' and 'trim axis group right'
- bugfixes for polar axes and log+stacked plots
- added style 'log ticks with fixed point'
- introduced patched tikz paths to simplify circles and ellipses within 
an axis
- patchplots lib: patch type=polygon
- some more bugfixes

1.5:
  - Contour plots,
  - Histograms,
  - Quiver plots,
  - patch plots (library)
       - Triangle Meshes
       - Bilinear Elements
       - Quadratic Triangles
       - Biquadratic Quadrilaterals
       - Coons Patches
  - Discrete colorbars,
  - Table sorting,
  - Linear regression,
  - Ternary diagrams,
       - Tieline Plots
  - Smith Charts
  - Polar axes,
  - Empty lines in input files result in interrupted plots,
  - PDF user defined coordinate mouse popups
  - CMYK colormaps and shadings,
  - new markers and cycle lists
  - access to axis limits,
  - \addplot3 graphics: pgfplots draws an appropriate axis for a 
three-dimensional(!) external png graphics
  - 3D axes: support to provide explicit unit vectors:
      - explicit unit vectors
     - explicit unit vectors which are uniformly rescaled to match 
width/height
  - 3D axes: improved support for unit vector ratios
  - improvements of the groupplot styles
  - preliminary support for (2d) bar plots in 3d axes
  - new shader 'faceted interp'
  - table package:
      - 'every nth row' style
     - 'comment chars' key to define comment characters in input files
     - 'skip first n' style
  - lots of smaller bugfixes (see ChangeLog for details)


1.4.1:
- improved compatibility to gnuplot 4.4

1.4:
Version 1.4 contains several new features, mostly work on details.
It fixes many bugs and provides the following improvements:
- detached legends
- detached colorbars
- ybar (and similar plots) can now be mixed with other plot types
   like line plots.
- improved legend formatting
- added 'restrict x to domain*' which cups coordinates outside of a 
specified domain (same for y and z)
- Added support for linear regression
- Inline tables,
- Lots of bug fixes

The next version will make a greater step when it is stable.


1.3.1:
Version 1.3.1 is a bugfix release containing
- improved parametric plots with gnuplot
- improved normalsize, small and footnotesize scale styles and added tiny
- a lot of bugfixes

1.3:
- improvements for two dimensional visualization, among them
     - axis equal,
     - color bars,
     - nodes near coords,
     - jumps in plots,
     - improved description positioning,
     - reverseable axis directions,
     - simpler alignment of adjacent axes,
     - units and a simplified user interface,
- new three dimensional line, scatter, mesh and surface plots,
- a copy of the automatic pdf externalization library,
- an improved manual enhanced with a lot of pdf cross references.

1.2.2:
- fixed a problem with the samples key,
- provides some smaller fixes and some manual improvements.
- added plot graphics.

1.2:
- completely rewritten math expression parser with extended data range,
- colormaps for scatter plots
- fine tuning for plot parameters.
- table package has been extended and is now a fully featured table 
typesetting, computing and postprocessing tool.

[Pgfplots-features] error bars: turn off the double style

[Ireneusz S. <ire...@gm...>]

Hi,

In my plots I use various line styles for plotting data, one of them 
is "double".  I plot error bars for the data points, and for the 
"double" line, the error bars are also plotted "double".  It's the way 
it's supposed to work, because the PGFPLOTS manual says:

 > Error bars inherit all drawing options of the
 > associated plot, but they use their own error
 > mark and additional style arguments.

I would like to have a regular error bar, i.e., solid and of the width 
of the plot lines, and so the following option works for all plots 
except the double one:

error bars/error bar style = {solid}

For the "double" lines, I managed to get it working with:

error bars/error bar style = {solid, double distance = -0.4 pt}

I wonder, though, whether there is a better way of turning off the 
double style for error bars.  I don't like setting manually the 
"double distance" value to get pretty error bars.


Best,
Irek

Re: [Pgfplots-features] \addplot [shift=...] breaks the legend box(?)

[Johannes H. <joh...@aa...>]

Dear Stefan,
	thanks for the reply and for the links. You're right, the shift is not
essential for producing a waterfall plot. I did eventually use 

\addplot ... y expr=\thisrowno{1} + <some y-shift value>...

for the desired outcome, so the issue is probably a case of mixing
incompatible features. 

My best,
 - J.

P.S. As my friend noted once, pgfplots gives too nice output - it will
ruin you for good, and now I weep every time I have to edit a *.doc
manuscript sent by a collaborator.

pe, 2016-10-07 kello 21:23 +0200, Stefan Pinnow kirjoitti:
> Dear Johannes,
> 
> > Dear all,
> > 	I have some data I want to plot as a waterfall plot by using the
> > shift={()} option in \addplot. Unfortunately this seems  to break the
> > legend entries/legend box in general.
> 
> For the record: As far as I know `shift' isn't supported by PGFPlots (at
> least I cannot find it in the manual) whereas `xshift' and `yshift' are.
> 
> Nonetheless, using these will "break" the legend, too. But I don't think
> that this is bug. We will see what Christian says about that.
> 
> ----------
> 
> Do you really need the shift key to produce the waterfall chart? Maybe these
> links provide some alternatives which avoid using one of the `shift' keys.
> 	<http://tex.stackexchange.com/questions/281988/waterfall-chart>
> 	<http://tex.stackexchange.com/questions/62699/waterfall-chart>
> 
> 
> Best regards,
> Stefan
> 
> ------------------------------------------------------------------------------
> Check out the vibrant tech community on one of the world's most 
> engaging tech sites, SlashDot.org! http://sdm.link/slashdot
> _______________________________________________
> Pgfplots-features mailing list
> Pgf...@li...
> https://lists.sourceforge.net/lists/listinfo/pgfplots-features

Re: [Pgfplots-features] \addplot [shift=...] breaks the legend box(?)

[Stefan P. <Mo...@gm...>]

Dear Johannes,

> Dear all,
> 	I have some data I want to plot as a waterfall plot by using the
> shift={()} option in \addplot. Unfortunately this seems  to break the
> legend entries/legend box in general.

For the record: As far as I know `shift' isn't supported by PGFPlots (at
least I cannot find it in the manual) whereas `xshift' and `yshift' are.

Nonetheless, using these will "break" the legend, too. But I don't think
that this is bug. We will see what Christian says about that.

----------

Do you really need the shift key to produce the waterfall chart? Maybe these
links provide some alternatives which avoid using one of the `shift' keys.
	<http://tex.stackexchange.com/questions/281988/waterfall-chart>
	<http://tex.stackexchange.com/questions/62699/waterfall-chart>


Best regards,
Stefan

[Pgfplots-features] \addplot [shift=...] breaks the legend box(?)

[Johannes H. <joh...@aa...>]

Dear all,
	I have some data I want to plot as a waterfall plot by using the
shift={()} option in \addplot. Unfortunately this seems  to break the
legend entries/legend box in general.

Here's a minimal example
----------------------------
\documentclass{article}
\usepackage{pgfplots}

\begin{document}
\begin{tikzpicture}
  \begin{axis}
    \addplot [samples=50] {x^2};
    \addplot [shift={(0,15)}, samples=50] {x^2};
    \addplot [shift={(0,25)}, samples=50] {x^2};
    \addlegendentry{A}; %
    \addlegendentry{B}; %
    \addlegendentry{C}; %
  \end{axis}
\end{tikzpicture}
\end{document}
---------------------------

With best wishes,
 - J.

[Pgfplots-features] pgfplots 1.14 is available

[Christian F. <cfe...@go...>]

Hi,

I am proud to announce the release of pgfplots 1.14 on CTAN and 
sourceforge. It will be available in TeX package manages within the next 
days.

Kind regards

Christian

1.14
- new feature: 'contour filled' (compat=1.14)
- new feature: building colormaps from other colormaps (see "Building 
Colormaps based on other Colormaps" in the manual)
- new feature: non-uniform colormaps (compat=1.14)
- new feature: colormaps defined on position of arbitrary magnitude
- new feature: colorbar as legend
- new feature: 'colorbar style={xtick=data}' positions tick labels at 
colormap positions
- fixed bug: pgfplots now handles incompatible changes of luatex
      loading \usepackage{pgfplots} _before_ pgf also allows makes older
      PGF versions compatible with luatex
- fixed bug: incompatibility between units + groupplots (bug 119)
- fixed bug: 'axis line shift' did not respect labels
- fixed bug: layers for axis lines were not respected
- fixed bug: two axes with fillbetween in the same picture failed due to 
clip paths on layers
- fixed bug: quiver plots with 'every arrow' failed to evaluate arrow 
tip length arguments
- fixed bug: \usepgfplotslibrary{colorbrewer}: colormap 'PuOr' was 
defined in reverse order

[Pgfplots-features] Problems with table/x error (index) key

[Ville K. <vil...@gm...>]

Hi!

I'm trying to implement certain pgfplots options into my own LaTeX 
package to shortcut writing papers. I am, however, running into problems 
when trying to set table/x error index key using \pgfplotsset. Here's an 
extract of the code:

\pgfplotsset{%

compat=1.13,

error bars/x dir=both,

error bars/x explicit,

error bars/y dir=both,

error bars/y explicit,

table/col sep=comma,

table/x index=0,

table/x error index=1,% this causes problems

table/y index=2,

table/y error index=3% and this as well}


The package works perfectly fine without the lines setting error column 
indices. However, when I try to run in a .tex file

\begin{tikzpicture}

\begin{axis}

\addplot table {data.csv};

\end{axis}

\end{tikzpicture}

it results in an error message "! Package pgfkeys Error: I do not know 
the key '/pgfplots/x error plus index', to which you passed '1', and I'm 
going to ignore it. Perhaps you misspelled it."

What is happening here? It seems to change the key given inside 
\pgfplotsset. Trying the complete key /pgfplots/table/x error index does 
not change the output.

Any help with this problem is highly appreciated. Thanks in advance!

Regards, Ville

Re: [Pgfplots-features] Version 1.13 has been released!

[Christian F. <cfe...@go...>]

Thanks for the praise, Nick!

I suppose citing pgfplots can be done using its web site (like 
http://pgfplots.sourceforge.net/).

Kind regards

Christian

Am 07.01.2016 18:30, schrieb Nick Papior:
> Great job Christian, your efforts are priceless!
>
> Do you have a citation for pgfplots such that your efforts, at least, 
> get some broader recognition?
>
> 2016-01-07 18:25 GMT+01:00 Christian Feuersaenger 
> <cfe...@go... <mailto:cfe...@go...>>:
>
>     Hi all,
>
>     I am proud to pronounce the release of pgfplots 1.13!
>
>     It is now available via CTAN http://www.ctan.org/pkg/pgfplots and
>     http://pgfplots.sourceforge.net/ and will become part of the
>     associated
>     TeX package managers (MikTeX, TexLive) within the next days.
>
>     A list of changes is attached to this email.
>
>     Kind regards
>
>     Christian
>
>     Changes:
>
>     1.13:
>     - fixed bug: incompatiblity between fillbetween and babel
>     - fixed bug: 'compat=1.9' (or newer) failed to work with log bar plots
>     - fixed bug: javascript incompatibility between Acrobat reader DC and
>     clickable lib
>     - fixed bug in polar axes: repaired default axis label positions
>     - fixed bug in polar axes: repaired support for sloped descriptions
>     - fixed bug: 'axis line style={draw=none}' had no effect unless
>     one had
>     2d boxed axes
>     - fixed bug: point meta expressions with '=' inside of them which
>     caused
>     compilation errors
>     - fixed bug: compat=1.12 combined with interrupted plots (by empty
>     line)
>     failed to work
>     - fixed bug: compat=1.12 combined with lualatex evaluated relational
>     operations (<, ==) with the wrong operator precedence
>     - fixed bug: fixed floor and ceil functions
>     - fixed bug: 'set layers' broke alignment features with different
>     manifestations
>             ('set layers' together with fill between in group plots,
>     together with at={} key, together with anchor)
>     - fixed bug: fill between and group plots: fill-between graphics was
>     missing (list of layers was lost)
>     - fixed bug: \addplot graphics {\newFile{P.pdf}}; failed to properly
>     expand the file name
>     - fixed bug: fill between + error bars resulted in compilation error
>     - fixed bug: stacked plots occasionally used a wrong coordinate
>     mapping
>     - fixed bug: provided suitable default legends for xbar and xbar
>     stacked
>     - fixed bug: 'dateplot' now computes missing 'date ZERO'
>     automatically.
>     'date ZERO' is managed automatically.
>     - new feature: Added support for search paths to find data files
>     (table/search path)
>     - new feature: 'xtick distance' allows to define tick positions by
>     distance
>     - new feature: "axis line shift=10pt" shifts axis lines together with
>     all their descriptions
>     - new feature: 'matrix plot' (alias 'imagesc') which allows cell-based
>     matrix plots
>     - new feature: unbounded point meta in mesh plots will be treated like
>     an invisible patch
>     - new feature: added high quality colormap 'viridis' (use 'colormap
>     name=viridis' to activate it)
>     - new feature: ensured that 'shader=flat' returns the same result for
>     every choice of 'z buffer' (starting with compat=1.13)
>     - new feature: utility function \pgfplotsforeachungrouped: added
>     support
>     for \breakforeach
>     - new feature: compat=1.13: activates sampling routine with less
>     rounding errors
>     - new feature: cycle multiindex list (uses the 'ith' element for every
>     provided sub-list)
>     - new feature: 'color of colormap=<value>' or 'color of
>     colormap=<value>
>     of <colormap name>'
>     - new feature: 'index of colormap=<value>' or 'index of
>     colormap=<value>
>     of <colormap name>'
>     - new feature: cycle list={[of colormap]}
>     - new feature: cycle list={[colors of colormap={0,100,500,100} of
>     viridis]}
>     - new feature: cycle list={[indices of colormap={0,1,2,3} of hot]}
>     - new feature: added support for 'stack negative=on previous|separate'
>     which allows separate plots for '+' and '-' (useful for bar plots)
>     - new feature: added support for 'colormap access=direct' combined
>     with
>     'shader=flat mean' and 'shader=interp':
>          it previously used to interpolate indices instead of colors
>     - new feature: \closedcycle now produces useful results for 3d
>     line plots
>     - new feature: added \usetikzlibrary{colorbrewer} and
>     \usepgfplotslibrary{colorbrewer}
>     - new feature: added 'colormap access=piecewise constant'
>     - new feature: added rudimentary support for "Filled Contours"
>     (based on
>     'colormap access=piecewise constant')
>     - new feature: added methods to build colormaps based on other
>     colormaps
>     using
>         "color of colormap", "index of colormap", "colors of colormap",
>     "samples of colormap", and "indices of colormap"
>
>     1.12.1: minor bugfix release
>     - fixed bug: incompatibility of 1.12 with unicode-math
>     - fixed bug: lua backend failed to work with table input and
>     dependency
>     on \coordindex
>     - fixed bug: lua backend failed with 'forget plot'
>
>     1.12: scalability + performance
>     - scalability: reduced TeX memory consumption for huge inline tables
>     - scalability: lualatex allows bigger 3d graphics
>     - scalability: lualatex now reduces compilation time for huge 3d
>     graphics
>         (requires \pgfplotsset{compat=newest} at the time of this writing)
>             surf shader=faceted: time reduced to 61% of pdflatex
>              (\addplot3[surf,samples=70] {exp(-x^2-y^2 +x*y)})
>
>          surf shader=interp: time reduced to 25% of pdflatex
>               (\addplot3[surf,shader=interp,samples=70] {exp(-x^2-y^2
>     +x*y)})
>
>          surf shader=interp, parametric: time reduced to 36% of pdflatex
>              (\addplot3[surf,shader=interp,z buffer=sort,
>     samples=50,domain=-1:0,y domain=0:2*pi] ({sqrt(1-x^2) * cos(deg(y))},
>     {sqrt( 1-x^2 ) * sin(deg(y))}, x);)
>
>          surf shader=faceted, parametric: time reduced to 40% of pdflatex
>              (\addplot3[surf,z buffer=sort, samples=50,domain=-1:0,y
>     domain=0:2*pi] ({sqrt(1-x^2) * cos(deg(y))}, {sqrt( 1-x^2 ) *
>     sin(deg(y))}, x);)
>
>     - scalability: lualated now reduces compilation time for boxplots.
>          10000 input points from table: time reduced to 23% of pdflatex
>     - fixed bug: 3d axes where all axes used log scale produced
>     compile failure
>     - fixed bug: axis equal did not respect "disabledatascaling"
>     - fixed bug: mesh legend was confused by color input=explicit
>     - fixed bug: colorbar style did not reset zmin/zmax
>     - fixed bug: memory allocation in FPU consumed too much memory
>     - fixed bug: boxplots: removed compilation error due to spurious
>     spaces
>     in options list
>     - fixed bug: boxplots: now uses standard quantile estimators: has
>     all 9
>     estimators of R and uses the same as in Excel as default (as of
>     compat=1.12)
>     - fixed bug: boxplots: sometimes failed to compute boxplot statistics
>     for special inputs
>     - fixed bug: boxplots: now, it sets 'mark=*' for outliers if there
>     is no
>     mark (compat=1.12)
>     - fixed regression: compat=1.11 broke arcs inside of axis
>     - new feature: ensured that arcs inside of axis make use of 'axis cs'
>     and 'axis direction cs' automatically
>     - new feature: ensured that (0:1) is evaluated with axis cs
>     - fixed bug: precision of transformation cartesian to polar was
>     too low
>     (added atan2 to fpu)
>     - fixed bug: z buffer caused 'mesh legend' to produce a compile error
>     - new feature: added 'y filter/.expression={y==3 ? nan : y}' to
>     simplify
>     coordinate filters
>
>
>
>     ------------------------------------------------------------------------------
>     _______________________________________________
>     Pgfplots-features mailing list
>     Pgf...@li...
>     <mailto:Pgf...@li...>
>     https://lists.sourceforge.net/lists/listinfo/pgfplots-features
>
>
>
>
> -- 
> Kind regards Nick