How to convert csv data with decimal csv

1 次查看（过去 30 天）

显示更早的评论

Mohammed Cha 2020-12-18

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/697230-how-to-convert-csv-data-with-decimal-csv

评论： Mohammed Cha 2020-12-23

采纳的回答： Walter Roberson

在 MATLAB Online 中打开

Hello Matlab Experts,

I am converting my csv (english) version to erman version. However getting the converted csv with extra space raw. Can anyone help me?

Thank you

clear all
str = fileread('C:\Users\Cha\Desktop\TOUCH.csv');
str = strrep(str, ',', ';');
str = strrep(str, '.', ',');
f = fopen('TOUCH.csv', 'wt');
fprintf(f, '%s', str);
fclose(f);

1 个评论
显示 -1更早的评论隐藏 -1更早的评论

Ive J 2020-12-18

Can you share your original file (TOUCH.csv) or a simulated one with the same structure?

请先登录，再进行评论。

请先登录，再回答此问题。

采纳的回答

Walter Roberson 2020-12-18

0
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/697230-how-to-convert-csv-data-with-decimal-csv#answer_579075

在 MATLAB Online 中打开

Change

f = fopen('TOUCH.csv', 'wt');

f = fopen('TOUCH.csv', 'w');

When you do fileread(), the character vector that is read in will have inside it any carriage returns and linefeeds that are in the original text. When you fwrite() when 'wt' is active, every newline to be written will be converted to carriage return followed by newline. So what was originally CR LF in your file would get written out as CR CR LF .

If you had reason to want to force CR LF line ending, then use 'w', but

str = strrep(strrep(str, char(13), ''), char(10), char([13 10]))

This would remove all existing carriage returns and then change all newline to carriage return followed by newline.

Or alternately you could write the entire replacement series with no strrep, using

str = regexprep(str, {',', '\.', '(?<!\r)\n'}, {';', ',', '\r\n'})

This does all of the replacements, and only converts \n to \r\n if the \n was not proceeded by a \r already.

8 个评论
显示 6更早的评论隐藏 6更早的评论

Walter Roberson 2020-12-23

在 MATLAB Online 中打开

myfilename = dir('*.csv');
allfilenames = {myfilename.name};
%get rid of ones that are already _ge.csv
mask = ~cellfun(@isempty, regexp(allfilenames, '_ge\.csv$', 'once'));
allfilenames(mask) = [];
numfiles= length(allfilenames);
for k = 1:numfiles
    oldname = allfilenames{K};
    newname = regexprep(oldname, '\.csv', '_ge.csv');
    mydata = fileread(oldname);
    str = regexprep(mydata, {',','\.','(?<!\r)\n'}, {';',',','\r\n'});
    f = fopen(newname, 'w');
    fwrite(f, str);
    fclose(f);
end