Textscan for comma seperated file with mixed format

Question

Thijs Van de Wiel 2021-1-20

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/722199-textscan-for-comma-seperated-file-with-mixed-format

评论： Thijs Van de Wiel 2021-1-20

I am trying to use the textscan function with a text file where the delimiter is a comma(,). I load the text file using:

f=fread(fid,'*char')';

Which gives me a variable f that looks like:

f = '2020-01-20 00:00:00,1,2,4,5,Not Connected,,,NaN,10,5';

The variable f contains strings and numeric data but I only want to use the numeric data. The format can change between measurements and the data does have multiple lines which I am not able to recreate now in the example.

Using code: textscan(f,'%f',11,'delimiter',',') will not give me an output as expected. I would expect a 11 by x cell array with only floating numbers or else a NaN.

How can I get this code to work?

0 个评论
显示 -2更早的评论隐藏 -2更早的评论

请先登录，再进行评论。

请先登录，再回答此问题。

Answer 1

Bob Thompson 2021-1-20

0
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/722199-textscan-for-comma-seperated-file-with-mixed-format#answer_602314

在 MATLAB Online 中打开

Have you tried using readtable? It has a delimited text option, and tends to be a bit better about handling different types of data.

f = readtable(fid,'Filetype','delimitedtext','Delimiter',',');

This is untested, because I don't have access to a sample file for you, but it should be something of that form.

1 个评论
显示 -1更早的评论隐藏 -1更早的评论

Thijs Van de Wiel 2021-1-20

Yes I have tried that, unfortuanetly it did take quite a lot of time to compute that for a large data set. Therefore I preferred using textscan since if the data set would be all floating numbers this method is the fastest.

请先登录，再进行评论。