How to create meshgrid for unequal metrix.

I am trying to plot pressure contour for roof (40'X125') where 8 main tap lines are given and another 7 sub tap lines starts after 99'; hence, tributary dimensions in 1 direction is reduced after 99'. Therefore, how to create meshgrid in this case. Image is given here for the reference.

Answer 1

[William Rose]

@Jigar,

The answer below shows how to create a grid that is non-uniform. The figures in the answer show how it worked.

https://www.mathworks.com/matlabcentral/answers/1818390-contour-plot-for-a-non-rectangular-object

However, the answer above does not reproduce a pattern like the one in your figure. meshgrid() cannot make a mesh like the one in your figure, because your figure has different numbers of rows for different columns.

I do not understand exactly what you are trying to do with the diagram shown. Do you have measurements at the dots shown on the diagram, and you want to make a contour plot using the measured values?

.

[Jigar]

Hello William,

First of all, thank you for your response. I am trying to create contour plot for points which are shown in the figure. Figure shows that Left part has only 8 lines of taps while right part shows 15 lines of taps which creates unequal matrix, and I want to make contour in a such a way that it captures pressure at all points.

[William Rose]

@Jigar,

I notice that the x-values for each column of dots are the same. That is, the columns of dots are all exactly aligned. The y-values , however, are uneven: the spacing in the y direction is half as wide for x>99.

If I had data at the dot points, and if it was my goal to make a contour plot, I would use meshgrid() to make a grid with 19x15 points. This grid will include all the points where I have data, plus some points where I do not have data. I would use linear interpolation to fill in the values at the points where I do not have data. After I have filled in the data at the grid points where there was no data, I would use contour() to make the contour plot.

[William Rose]

@Jigar,

[X,Y]=meshgrid([0:12.5:75,81:6:99,102:26/7:128],[1:38/14:39]);
plot(X,Y,'r+'); xlim([0,128])
z=zeros(size(X));
for i=2:2:14
    z(i,1:10)=1;        %z=1 where we are missing data
end
z(15,4)=1; z(5,7)=1;    %z=1 missing data points, special cases
zc=~z;                  %zc=1 where we DO have data
hold on; plot(X(zc),Y(zc),'b.')

The + symbols are all the points in the mesh. The blue points are where we have data. I think the blue points above match your blue dots. The matrix z equals 1, or true, where you need to interpolate a value.

[Jigar]

@William Rose,

Noted and thank you for your response. I will try this.

[William Rose]

@Jigar, you're welcome. Good luck with your work.