Reading a large binary file

60 ビュー (過去 30 日間)
Dominik Rhiem
Dominik Rhiem 2023 年 10 月 18 日
コメント済み: Dominik Rhiem 2023 年 10 月 19 日
I want to load a 8 GB binary data file into Matlab. My computer has 24 GB of RAM. It runs out of memory, despite me closing literally everything besides Matlab, and I don't understand why. The file should not take more than its size, right?
Code for loading is simply:
fd = fopen(['data.bin'], 'r');
data = fread(fd, 'int16');
fclose(fd);

採用された回答

James Tursa
James Tursa 2023 年 10 月 18 日
編集済み: James Tursa 2023 年 10 月 18 日
Try
data = fread(fd, '*int16');
When you use 'int16' for the type to read, MATLAB reads as int16 and then converts to double (four times the memory). When you use '*int16', MATLAB keeps the type as int16. E.g.,
% Create a small sample file
fid = fopen('junk.bin','wb');
fwrite(fid,int16(4),'int16');
fclose(fid);
% read it with 'int16'
fid = fopen('junk.bin','rb');
x = fread(fid,'int16')
x = 4
class(x) % double, not what you wanted
ans = 'double'
% read it with '*int16'
frewind(fid);
x = fread(fid,'*int16')
x = int16 4
class(x) % int16, this is what you wanted
ans = 'int16'
fclose(fid);
  1 件のコメント
Dominik Rhiem
Dominik Rhiem 2023 年 10 月 19 日
Positive side effect: This is extremely fast.

サインインしてコメントする。

その他の回答 (0 件)

カテゴリ

Help Center および File ExchangeLarge Files and Big Data についてさらに検索

タグ

Community Treasure Hunt

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

Start Hunting!

Translated by