Out of memory error for big data

3 次查看(过去 30 天)
Sharif Khalil
Sharif Khalil 2020-12-29
The masterDataMatrxOffset, contains 464780 rows and 256 columns of complex numbers. I use meshgrid to get 256X256 matrix of the real and imaginary and store it in a cell, but I get this error: Is there a way to overcome the out of memory?
Out of memory.
Error in meshgrid (line 62)
yy = repmat(ycol,size(xrow));
Error in auto_encoder (line 34)
[Po,Ph] = meshgrid(power(obsrv,:),phase(obsrv,:));
More information
tic;
clear; clc; close all;
%% Load Files
load masterDataMatrxOffset;
%% Extract Parameters
iAntenna = size(masterData,2) - 16;
data = masterData(:,1:iAntenna);
power = normalize(real(data),'range');
phase = normalize(imag(data),'range');
Region = masterData(:,261);
iReceiver = length(data);
iRegion = max(Region);
%% Initializations
xTrn = cell(1,max(Region));
xTst = cell(1,max(Region));
datamat = cell(1,size(data,1));
%% Split into training and testing
for idx = 1:iRegion
IDX = find(Region==idx);
train = sort(randperm(size(IDX,1),ceil(0.8*size(IDX,1))));
trnidx = IDX(train);
tstidx = setdiff(IDX,trnidx);
xTrn{idx} = trnidx;
xTst{idx} = tstidx;
end
xTrnidx = cell2mat(xTrn');
xTstidx = cell2mat(xTst');
Trainreg = zeros(numel(Region), max(Region));
Trainreg(sub2ind(size(Trainreg), 1:numel(Region), Region')) = 1;
tTrain = Trainreg';
for obsrv = 1:size(power,1)
sprintf('observation = %d',obsrv);
[Po,Ph] = meshgrid(power(obsrv,:),phase(obsrv,:));
datamat{obsrv} = Po.^2 + Ph.^2;
end
toc;
  4 个评论
Ive J
Ive J 2020-12-29
编辑:Ive J 2020-12-29
You still can store them in a cell array
myCell = {};
myCell{1} = tall(1:100); myCell{2} = tall(101:200);
myCell{3} = myCell{1}.^myCell{2};
myCell =
1×3 cell array
{1×100 tall} {1×100 tall} {1×100 tall}
Sharif Khalil
Sharif Khalil 2020-12-31
编辑:Sharif Khalil 2020-12-31
It still shows out of memory. The cell is 1X464780, but the matrix inside each cell is 256X256.
for obsrv = 1:size(power,1)
[Po,Ph] = meshgrid(power(obsrv,:),phase(obsrv,:));
Talldata = tall(Po.^2 + Ph.^2);
datamat{obsrv} = Talldata;
end

请先登录,再进行评论。

回答(0 个)

类别

Help CenterFile Exchange 中查找有关 Large Files and Big Data 的更多信息

标签

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by