[Tex/LaTex] pgfplots: How to imitate MATLAB filled contour plot

pgfplots

Following this answer, I have a data file P.dat which I need to plot it as a filled contour plot.

However, I got this error

Package pgfplots Error: CRITICAL: shader=interp: got unsupported pdf
shading type '0'. This may corrupt your pdf!. \end{axis}

I would be grateful if I could know the error source and how I can make the output match the MATLAB desired one.

P.dat

http://pastebin.com/PAvQMxdC

MWE

\RequirePackage{luatex85}
\documentclass[tikz]{standalone}

\usepackage{pgfplots}
\pgfplotsset{compat=newest}

\begin{document}
    \begin{tikzpicture}
    \begin{axis}
    \addplot3[contour filled] table {P.dat};
    \end{axis}
    \end{tikzpicture}
\end{document}

MATLAB Desired Output

Update 1

I made another data file with NaN z values to account for the void data (white space) and specified the number of rows and columns, but I got this undesired output.

P2.dat

http://pastebin.com/CCUjeQUm

MWE 2

\RequirePackage{luatex85}
\documentclass{standalone}

\usepackage{pgfplots}
\usepgfplotslibrary{patchplots}
\pgfplotsset{compat=newest}

\begin{document}
    \begin{tikzpicture}
    \begin{axis}[view={0}{90}]
    \addplot3[contour filled,mesh/rows=31,mesh/cols=11,mesh/check=false] table {P2.dat};
    \end{axis}
    \end{tikzpicture}
\end{document}

Output 2

Update 2

Recalling my raw MATLAB data, how can I remove all the points whose z values equal to or exceed 1723 in order to get an output similar to my desired one?

P3.dat

http://pastebin.com/jvNVfJMU

MWE 3

\RequirePackage{luatex85}
\documentclass{standalone}

\usepackage{pgfplots}
\usepgfplotslibrary{patchplots}
\pgfplotsset{compat=newest}

\begin{document}
    \begin{tikzpicture}
    \begin{axis}[view={0}{90},colorbar, point meta max=1723, point meta min=300,]
    \addplot3[contour filled={number = 25,labels={false}},mesh/rows=31,mesh/cols=11,mesh/check=false
    ] table {P3.dat};
    \end{axis}
    \end{tikzpicture}
\end{document}

Output 3

Best Answer

Here I present two solutions.

Solution 1 (left part of the image)

This tries to reproduce the Matlab figure with the capabilities of PGFPlots. To "confirm" that I did it right, I first saved your Matlab image and cropped the axis parts. Then I added this as \addplot graphics and on top of that I added the real plot, i.e. \addplot contour filled plot in 50% transparent. That allowed to check, if I found the interval boundaries right.

Said that I think your above statement is wrong. It seems that you removed the color for all values >1600. This also makes sense, because the maximum value in the P3.dat file is 1723 ...

Solution 2 (right part of the image)

Here I just used the above cropped Matlab image and reproduced the colorbar.

Comparison

As you can see in solution 1 there are some "artifacts" which don't make the result as smooth as the Matlab result. That is, because the contours calculation/visualization only depends on the features of your PDF viewer. Said that it could be that your result differs from mine. I made the screenshot from a view in Acrobat Reader XI.

That is why I favor solution 2.

To improve the result you should modify your Matlab view to only show the contours, that means remove the axis and the grid lines. Then the only difference could be the colors used/shown in Matlab contour plot and the colorbar calculated by PGFPlots. Concrete I mean that one could use RGB colors and the other CMYK. But since you have the illustrator as you said, you could check for that and adapt one of the both parts, i.e. the Matlab or PGFPlots output.

You could also create a "pure", that means also without any axis, version of the colorbar in Matlab and also import this graphics into the PGFPlots colorbar. Of course then colors are identical.

For more details on how the solutions work, please have a look at the comments in the code.

% used PGFPlots v1.14
\RequirePackage{luatex85}
\documentclass{standalone}
\usepackage{pgfplots}
\pgfplotsset{
    % you need at least this `compat' level or higher to use the below features
    compat=1.14,
    % define a "white" colormap for the white part of the image
    colormap={no data}{
        color=(white)
%        color=(white)
        color=(red)
    },
    % load this colormap which is later used
    colormap/bluered,
    % define the "parula" colormap that was used to create the exported image
    % from Matlab
    % (borrowed from http://tex.stackexchange.com/a/336647/95441)
    colormap={parula}{
        rgb255=(53,42,135)
        rgb255=(15,92,221)
        rgb255=(18,125,216)
        rgb255=(7,156,207)
        rgb255=(21,177,180)
        rgb255=(89,189,140)
        rgb255=(165,190,107)
        rgb255=(225,185,82)
        rgb255=(252,206,46)
        rgb255=(249,251,14)
    },
}
\begin{document}
    \begin{tikzpicture}
        \begin{axis}[
            view={0}{90},
            colorbar,
            % modify the style of the colorbar a bit
            colorbar style={
                ytick distance=200,
                ymax=1600,
            },
            % this key--value is needed because of the `\addplot graphics'
            enlargelimits=false,
        ]
            % import the "exported" graphics
            \addplot graphics [
                xmin=0,
                xmax=50,
                ymin=0,
                ymax=150,
            ] {P3};
            % now try to reproduce the style of the exported graphics
            \addplot3 [
                % for that use, e.g., the `countour filled' feature ...
                contour filled={
                    % ... in combination with the `levels of colormap' feature
                    % which allows to customize the used colormap
                    levels from colormap={
                        % this part of the colormap is for the "colored" part
                        of colormap={
                            % here we use the above initialized `bluered' colormap
                            bluered,
%                            % (`viridis' is a colormap which is similar to the
%                            %  used `parula' comormap in Matlab.
%                            %  But because the yellow is hard to identify
%                            %  in this context we use the above colormap)
%                            viridis,
                            % with this we state there is more to come
                            target pos max=,
                            % and here we state where the corresponding levels
                            % should *start*
                            target pos={200,400,600,800,1000,1200,1400},
                        },
                        % here comes the second part of the colormap which
                        % should have no color which isn't possible or at least
                        % I don't have an idea how to do it ...
                        of colormap={
                            % ... so I use a "white" colormap instead
                            no data,
                            % here the lower end isn't needed because that was
                            % specified in the first part of the colormap
                            target pos min=,
                            % and here is the corresponding interval *start*
                            % for that colormap
                            % (as you can see -- or not ;) -- the white starts
                            %  at position/values >=1600)
                            target pos={1600},
                        },
                    },
                },
                % you need only to provide `rows' or `cols' because
                % PGFPlots can then calculate the other value together with
                % the provded number of data points
                mesh/rows=31,
                % make the plot half transparent to check that the `target pos'
                % of the colormap are chosen correct
                opacity=0.5,
            ] table {P3.dat};
        \end{axis}
    \end{tikzpicture}
    \begin{tikzpicture}
        \begin{axis}[
            % show the colorbar
            colorbar,
            % because there is no real plot where PGFPlots can get the `meta'
            % data from, one has to provide them manually
            point meta min=200,
            point meta max=1800,
            %%% here we define the needed colormap and its style again
            % we want to use constant intervals in the colormap
            colormap access=piecewise const,
            % and also here we have to provide the limits again ...
            of colormap/target pos min*=200,
            of colormap/target pos max*=1800,
            % ... and use this feature which makes easier to provide the
            % samples at the right position
            % (please have a look at the PGFPlots manual for more details)
            of colormap/sample for=const,
            % this is similar to the above example
            colormap={CM}{
                of colormap={
                    % ... except that we use the `parula' colormap here
                    parula,
                    target pos max=,
                    target pos={200,400,600,800,1000,1200,1400,1600},
                },
                of colormap={
                    no data,
                    target pos min=,
                    % here you can use an arbitrary value which is greater than
                    % the last `target pos' of the previous colormap part of course.
                    % But here I tried to "overwrite" the last color of the
                    % colormap, i.e. the "bright" yellow, as well by just
                    % giving it a very small interval
                    % (to show the effect increase the `ymax' value in the
                    %  `colorbar style')
                    target pos={1601},
                },
            },
            % modify the style of the colorbar a bit
            colorbar style={
                ytick distance=200,
                ymin=300,
                ymax=1600,
            },
            % this key--value is needed because of the `\addplot graphics'
            enlargelimits=false,
        ]
            \addplot graphics [
                xmin=0,
                xmax=50,
                ymin=0,
                ymax=150,
            ] {P3};
        \end{axis}
    \end{tikzpicture}
\end{document}

Related Solutions

[Tex/LaTex] plotting two time series with bounds

This happens because PGFPlots only uses one "stack" per axis: You're stacking the second confidence interval on top of the first. The easiest way to fix this is probably to use the approach described in "Is there an easy way of using line thickness as error indicator in a plot?": After plotting the first confidence interval, stack the upper bound on top again, using stack dir=minus. That way, the stack will be reset to zero, and you can draw the second confidence interval in the same fashion as the first:

\documentclass{standalone}
\usepackage{pgfplots, tikz}

\usepackage{pgfplotstable}

\pgfplotstableread{
temps   y_h y_h__inf    y_h__sup    y_f y_f__inf    y_f__sup    

1   0.237340    0.135170    0.339511    0.237653    0.135482    0.339823    
2   0.561320    0.422007    0.700633    0.165871    0.026558    0.305184    
3   0.694760    0.534205    0.855314    0.074856    -0.085698   0.235411    
4   0.728306    0.560179    0.896432    0.003361    -0.164765   0.171487    
5   0.711710    0.544944    0.878477    -0.044582   -0.211349   0.122184    
6   0.671241    0.511191    0.831291    -0.073347   -0.233397   0.086703    
7   0.621177    0.471219    0.771135    -0.088418   -0.238376   0.061540    
8   0.569354    0.431826    0.706882    -0.094382   -0.231910   0.043146    
9   0.519973    0.396571    0.643376    -0.094619   -0.218022   0.028783    
10  0.475121    0.366990    0.583251    -0.091467   -0.199598   0.016664    
}{\table}

\begin{document}
\begin{tikzpicture}

    \begin{axis}
    % y_h confidence interval
    \addplot [stack plots=y, fill=none, draw=none, forget plot]   table [x=temps, y=y_h__inf]   {\table} \closedcycle;
    \addplot [stack plots=y, fill=gray!50, opacity=0.4, draw opacity=0, area legend]   table [x=temps, y expr=\thisrow{y_h__sup}-\thisrow{y_h__inf}]   {\table} \closedcycle;
    % subtract the upper bound so our stack is back at zero
    \addplot [stack plots=y, stack dir=minus, forget plot, draw=none] table [x=temps, y=y_h__sup] {\table};

    % y_f confidence interval
    \addplot [stack plots=y, fill=none, draw=none, forget plot]   table [x=temps, y=y_f__inf]   {\table} \closedcycle;
    \addplot [stack plots=y, fill=gray!50, opacity=0.4, draw opacity=0, area legend]   table [x=temps, y expr=\thisrow{y_f__sup}-\thisrow{y_f__inf}]   {\table} \closedcycle;

    % the line plots (y_h and y_f)    
    \addplot [stack plots=false, very thick,smooth,blue]  table [x=temps, y=y_h]   {\table};
    \addplot [stack plots=false, very thick,smooth,blue]  table [x=temps, y=y_f]   {\table};
    \end{axis}

\end{tikzpicture}
\end{document}

[Tex/LaTex] Polar Plots – layout and extraction of crucial parameters

An experiment (30% solved, 70% remaining...)

I have rearranged your data from Matlab format manually (with and without the third column which may store the level), but I am setting color per plot/contour manually because I don't know how to extract a color from the used colormap or shading at a specific location, yet.

Notice: The format is that it contains the level and number of points per one contour, then angle and distance of that point, one point per line, and again, the level and number of points for another contour etc.

That's probably for a separate question at TeX.SX as that color is inbetween two known colors with a known distance from each other. I was thinking of my own color map to handle it and then it would be easy to match the colors at those levels back, but I am not for it right now. Close topic to this is PGFPlots: synchronize the filling of the bars with the colorbar.

I am enclosing what I have got right now. Please notice that gray color doesn't fit to the original colorbar where red color is used at top, I did that on purpose to highlight a new arisen problem in this approach.

\documentclass[margin=10]{standalone}
\usepackage{pgfplots}
\usepgfplotslibrary{polar}
%\pgfplotsset{compat=newest} % 1.10

\begin{document}
%\ifx\relax % An original version...
\begin{tikzpicture}
\begin{polaraxis}[
  title=\texttt{Station 1 (quantity [units])}, colorbar,
  ymax=.15, grid=major, xmin=30, yticklabels={0,,0.1,,0.2},
  %xlabel={\texttt{quantity (units)}$\rightarrow$},
  ]
\addplot[contour prepared={labels=false}, contour prepared format=matlab] file {mal-polar-data.txt};
\end{polaraxis}
\end{tikzpicture}
%\fi% End of an original version...
\hspace{8mm}%
\begin{tikzpicture}
\begin{polaraxis}[title=\texttt{Station 1 (quantity [units])}, colorbar, ymax=.17, 
 grid=major, xmin=30, yticklabels={0,,0.1,,0.2}, 
%xlabel={\texttt{quantity (units)}$\rightarrow$}
]
\addplot[mark=none, fill=blue] table { % contour prepared, color is 0.006, points 35
99.347  0.062 0.006
120.000 0.059 0.006
150.000 0.056 0.006
180.000 0.055 0.006
210.000 0.055 0.006
240.000 0.056 0.006
270.000 0.059 0.006
290.917 0.062 0.006
300.000 0.064 0.006
320.125 0.071 0.006
330.000 0.077 0.006
337.057 0.081 0.006
346.617 0.092 0.006
347.017 0.106 0.006
332.627 0.120 0.006
330.000 0.122 0.006
300.000 0.136 0.006
291.482 0.138 0.006
270.000 0.144 0.006
240.000 0.148 0.006
210.000 0.150 0.006
180.000 0.150 0.006
150.000 0.148 0.006
120.000 0.144 0.006
98.788  0.138 0.006
90.000  0.136 0.006
60.000  0.122 0.006
57.587  0.120 0.006
43.064  0.106 0.006
43.467  0.092 0.006
53.117  0.081 0.006
60.000  0.077 0.006
70.094  0.071 0.006
90.000  0.064 0.006
99.347  0.062  0.006
};
\addplot[mark=none, fill=blue!30] table { % color 0.012,   points 31
165.383 0.062
180.000 0.061
210.000 0.061
224.441 0.062
240.000 0.063
270.000 0.066
289.722 0.071
300.000 0.074
316.253 0.081
327.294 0.092
327.761 0.106
311.104 0.120
300.000 0.126
270.000 0.134
240.000 0.137
226.737 0.138
210.000 0.139
180.000 0.139
163.077 0.138
150.000 0.137
120.000 0.134
90.000  0.126
79.189  0.120
62.394  0.106
62.865  0.092
73.998  0.081
90.000  0.074
100.535 0.071
120.000 0.066
150.000 0.063
165.383 0.062 
};
\addplot[mark=none, fill=yellow] table { % color 0.018,   points 27
129.729 0.071
150.000 0.068
180.000 0.066
210.000 0.066
240.000 0.068
260.302 0.071
270.000 0.072
298.620 0.081
300.000 0.082
315.003 0.092
315.702 0.106
300.000 0.117
291.759 0.120
270.000 0.128
240.000 0.132
210.000 0.134
180.000 0.134
150.000 0.132
120.000 0.128
98.513  0.120
90.000  0.116
74.552  0.106
75.257  0.092
90.000  0.082
91.703  0.081
120.000 0.072
129.729 0.071 
};
\addplot[mark=none, fill=orange] table { % color 0.024,   points 27
170.180 0.071
180.000 0.070
210.000 0.070
219.672 0.071
240.000 0.072
270.000 0.077
282.910 0.081
300.000 0.090
302.713 0.092
303.644 0.106
300.000 0.108
273.691 0.120
270.000 0.122
240.000 0.128
210.000 0.130
180.000 0.130
150.000 0.128
120.000 0.122
116.443 0.120
90.000  0.108
86.710  0.106
87.649  0.092
90.000  0.091
107.293 0.081
120.000 0.077
150.000 0.072
170.180 0.071 
};
\addplot[mark=none, fill=orange!50] table { % color 0.030,   points 21
123.324 0.081
150.000 0.076
180.000 0.074
210.000 0.074
240.000 0.076
266.766 0.081
270.000 0.082
291.525 0.092
292.553 0.106
270.000 0.116
253.470 0.120
240.000 0.123
210.000 0.126
180.000 0.126
150.000 0.123
136.499 0.120
120.000 0.116
97.724  0.106
98.744  0.092
120.000 0.082
123.324 0.081 
};
\addplot[mark=none, fill=red] table { % color 0.036   points 21
141.306 0.081
150.000 0.079
180.000 0.077
210.000 0.077
240.000 0.079
248.619 0.081
270.000 0.087
280.650 0.092
281.881 0.106
270.000 0.111
240.000 0.119
227.611 0.120
210.000 0.122
180.000 0.122
162.193 0.120
150.000 0.119
120.000 0.111
108.314 0.106
109.536 0.092
120.000 0.087
141.306 0.081 
};
\addplot[mark=none, fill=red!50] table { % color 0.043   points 19
165.937 0.081
180.000 0.080
210.000 0.080
223.888 0.081
240.000 0.083
269.740 0.092
270.000 0.094
271.210 0.106
270.000 0.106
240.000 0.115
210.000 0.119
180.000 0.119
150.000 0.115
120.000 0.106
118.903 0.106
120.000 0.095
120.377 0.092
150.000 0.083
165.937 0.081 
};
\addplot[mark=none, fill=red!20] table { % color 0.049   points 13
132.787 0.092
150.000 0.087
180.000 0.083
210.000 0.083
240.000 0.087
257.216 0.092
259.102 0.106
240.000 0.111
210.000 0.115
180.000 0.115
150.000 0.111
130.918 0.106
132.787 0.092 
};
\addplot[mark=none, fill=gray] table { % color 0.055   points 13
145.197 0.092
150.000 0.091
180.000 0.086
210.000 0.086
240.000 0.091
244.692 0.092
246.810 0.106
240.000 0.108
210.000 0.112
180.000 0.112
150.000 0.108
143.098 0.106
145.197 0.092
};
\addplot[mark=none, fill=gray!50] table { % color 0.061   points 9
163.076 0.092
180.000 0.090
210.000 0.090
226.734 0.092
230.718 0.106
210.000 0.109
180.000 0.109
159.069 0.106
163.076 0.092 
};
\end{polaraxis}
\end{tikzpicture}

\end{document}

mwe