MATLAB Examples

Example

Introduction

```The example I use here is inspired from
http://rucool.marine.rutgers.edu/manuals/tag/pcolor/.
First, we generate coordinates as a [3000 x 1] vectors. If we have to
create a pseudocolor plot of the corresponding grid, this is goind to be
time consumming, because the matrix will be [3000 x 3000]. By using the
function scatter2Pcolor and a limited resolution for the grid, we speed
up the computation time, and still have a good result.```
```clear all;close all;clc; % generate coordinates x = rand(3000,1)*200-100; %Lon coordinates y = rand(3000,1)*200-100; %Lat coordinates z = 10*sin((x*pi/180).^2+(y*pi/180).^3)+20; %Faux-measured data ```

Minimal code given by rucool.marine.rutgers.edu :

```tic [X,Y] = meshgrid(unique(x),unique(y)); Z = griddata(x,y,z,X,Y); figure pcolor(X,Y,Z); shading flat; colorbar; hold on; plot(x,y,'k.') %Original data points; toc ```
```Elapsed time is 14.445450 seconds. ```

Application of function scatter2Pcolor without options

```tic figure scatter2Pcolor(x,y,z) toc ```
```Elapsed time is 0.263232 seconds. ```

Application of function scatter2Pcolor with multiple options

```tic figure scatter2Pcolor(x,y,z,'display','yes','resolution',200,'Xlabel',... 'my x label','Zlabel','my Z label','Ylabel','my Y label'); toc ```
```Elapsed time is 0.222991 seconds. ```

Case with nan or inf values for the coordinates x and y

```y(1:10:end)=nan; x(5:10:end)=inf; tic figure scatter2Pcolor(x,y,z,'display','yes','resolution',200,'Xlabel',... 'my x label','Zlabel','my Z label','Ylabel','my Y label'); toc ```
```Warning: The coordinates x of the input points contains Inf values. Warning: The coordinates y of the input points contains NaN values. Elapsed time is 0.217517 seconds. ```

Conclusions

`The function scatter2Pcolor is found to be particularly useful when:`
• the set of data is heterogeneously distributed in the grid
• The size of the data set is large, leading to a high computational cost, if pcolor is directly used.