[Tex/LaTex] Logarithmic y-axis and restrict y to domain fails

pgfplots

I'm trying to plot data with a logarithmic y-axis and restricting the y-data to a specified domain, but compilation fails for a reason I don't understand. This is the code:

\documentclass{standalone}
\usepackage{pgfplots}
\begin{document}%
\begin{tikzpicture}
\begin{axis}[ymode=log, restrict y to domain=1:6]
\addplot coordinates {(0, 1.1e1) (1, 1.7e3) (2, 0.0e0) (2, 3.4e7) (4, 8.1e5)};
\end{axis}
\end{tikzpicture}
\end{document}

Running it results in an error message:

! Missing number, treated as zero.
<to be read again> 
                   i
l.6 ..., 1.7e3) (2, 0.0e0) (2, 3.4e7) (4, 8.1e5)};

According to the pgfplots manual for version 1.5.1 (29 December 2011), page 272 (about restrict x/y/z to domain), For logarithmic axes, min and max are logs of the respective values. If I disable either the logarithmic y-axis or the domain-restriction, the code runs successfully and output is as expected (but not as desired).

Software versions:

pdfTeX, Version 3.1415926-2.3-1.40.12 (TeX Live 2011) (format=pdflatex 2012.5.30)
pgfplots 2011/12/29 v1.5.1 (git show 1.5.1-4-g53e640f )
tikz 2010/10/13 v2.10 (rcs-revision 1.76).

What's going on? What am I doing wrong? Or is this a bug?

Best Answer

This is not a direct answer however it might help to get what you want to do. I also don't know why it doesn't work.

\documentclass{standalone}
\usepackage{pgfplots}
\begin{document}%
\begin{tikzpicture}
\begin{semilogyaxis}[restrict expr to domain={y}{10:1e6},unbounded coords=discard]
\addplot coordinates {(0, 1.1e1) (1, 1.7e3) (2, 1.0e0) (2, 3.4e7) (4, 8.1e5)};
\end{semilogyaxis}
\end{tikzpicture}
\end{document}

If you remove the unbounded coords=discard it is set to jump.

enter image description here

Related Solutions

[Tex/LaTex] Single double-logarithmic axis

As Christian Feuersänger said, you can use a y coord trafo to transform the coordinates on the fly. The tick labels would usually be re-transformed using y coord inv trafo, but the precision of the math engine isn't high enough for this (1000 becomes 997.8), so you'll have to provide the labels explicitly:

\documentclass{article}
\usepackage{pgfplots}
\begin{document}


\begin{tikzpicture}
\begin{axis}[
    y coord trafo/.code=\pgfmathparse{log10(log10(#1))},
    domain=0:2,
    ymax=10000,
    ytick={10,100,1000,10000},
    yticklabels={10,100,1000,10000},
    extra y ticks={2,...,9,20,30,...,90,200,300,...,900,2000,3000,...,9000},
    extra y tick labels={},
    every extra y tick/.style={major tick length=3pt}
]
\addplot {exp(exp(x))};
\end{axis}
\end{tikzpicture}
\end{document}

[Tex/LaTex] Opposite command of “restrict y to domain”

This is what y filter is for (it's what restrict y to domain uses internally as well). Within y filter/.code, you have access to the current y coordinate and can run tests on it. At the end of the code, the current content of \pgfmathresult is used as the new y coordinate.

In your case, you can use a filter like

y filter/.code={\pgfmathparse{y / or(y<3, y>10)}}

to restrict the range to points less than 5 or greater than 10. The or(...) statement will evaluate to zero otherwise, leading to a division by zero, which causes the points to quietly be discarded.

You can wrap this in a convenience style, of course, like in this example:

\documentclass{article}
\usepackage{pgfplots}

\pgfplotstableread{
1 1
2 2
3 4
4 8
5 16
}\datatable

\pgfplotsset{
    ignore points in x range/.style args={#1:#2}{
        x filter/.code={\pgfmathparse{x/or(x<#1, x>#2)}}
    },
    ignore points in y range/.style args={#1:#2}{
        y filter/.code={\pgfmathparse{y/or(y<#1, y>#2)}}
    }
} 

\begin{document}
\begin{tikzpicture}
\begin{axis}
\addplot [only marks, gray, restrict y to domain=3:10] table {\datatable};
\addplot [only marks, red, ignore points in y range=3:10] table {\datatable};
\end{axis}
\end{tikzpicture}
\end{document}

Best Answer

Related Solutions

[Tex/LaTex] Single double-logarithmic axis

[Tex/LaTex] Opposite command of “restrict y to domain”

Related Question