Adjust randomly-distributed matrix to a organised matrix

1 次查看（过去 30 天）

Maurício Andrade 2022-10-11

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/1822608-adjust-randomly-distributed-matrix-to-a-organised-matrix

评论： Mathieu NOE 2022-10-12

Hello everyone,

I have an adjustment of matrix to handle and I would like to check for help here.

I have a matrix of temperature (T) along a LAT-LON grid, being T(57, 675780) and LAT-LON (675780, 1). I removed duplicates of LAT and LON, but I would like to generate a matrix of T 2D (I mean, considering T(1,:) as the surface layer of the ocean) adjusted to a new LAT (674, 1) and LONG (1440, 1) to plot sea surface temperature. The file is not allowed to upload here since it exceeds 5MB. :(

So, does anyone have a tip to try to solve this? Thank you very much :D

3 个评论
显示 1更早的评论隐藏 1更早的评论

Maurício Andrade 2022-10-11

test.mat

I tried to get a piece of the data. So, I added it here as an example. Lat, Lon and temp as (675780, 1). I used these steps on LAT and LON to remove duplicates and create a grid:

[lat, idx] = sort(lat);

lon = lon(idx);

temp = temp(idx);

[lat,ia,ib] = unique(lat); % lat --> (674, 1)

[lon, ja, jb] = unique(lon); %lon --> (1440, 1)

The step I would like to learn is to create a matrix for temp (2D) to be adjusted to the adjusted lat and lon.

Mathieu NOE 2022-10-12

在 MATLAB Online 中打开

hello

I am not sure about what you are trying to achieve.

your code will remove 99.8 % of the available data and you get almost nothing left from your map

below I tried to show some alternative if the goal is to "downsample" your map

there is no need to create a 2D temp array as scatter works with vectors .

clc
clearvars
load('test.mat')
% this is not needed as lat is monotonically increasing 
% [lat, idx] = sort(lat);
% long = long(idx);
% temp = temp(idx);
% let's plot the data as they come
figure(1);
pix = 3;  % area of each marker (in points^2).
scatter(long,lat,pix,temp,'filled')
% "decimated" display (any use ?)
decim = 10; % decimation factor
nn = numel(long);
new_ind = 1:decim:nn;
new_lat = lat(new_ind);
new_long = long(new_ind);
new_temp = temp(new_ind);
figure(2);
scatter(new_long,new_lat,pix*decim,new_temp,'filled')
% % your code
% [lat,ia,ib] = unique(lat); % lat --> (674, 1)
% long = long(ia); % lon --> (674, 1)
% temp = temp(ia);
% your code / alternative
[long,ia,ib] = unique(long); % long --> (1440, 1)
lat = lat(ia); % lat --> (1440, 1)
temp = temp(ia);
figure(3);
scatter(long,lat,5,temp,'filled')