Reading large CSV files

Question

Sanchit Sharma 2022-4-2

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/1686674-reading-large-csv-files

回答： Esha Chakraborty 2022-4-5

Hello,

I have a 28GB Large .csv file that I am trying to read. I have tried readmatrix() and readtable(). Both functions are giving me below error:

Caught "std::exception" Exception message is:

Failed to convert character code.

Could you please provide me a solution.

Thanks

3 个评论
显示 1更早的评论隐藏 1更早的评论

Sanchit Sharma 2022-4-2

编辑：Sanchit Sharma 2022-4-2

yes i have 512GB ram

per isakson 2022-4-2

编辑：per isakson 2022-4-2

在 MATLAB Online 中打开

Propably, your file contains some strange characters, e.g. to indicate missing data. The error message indicates that. One way to spot the position in the file that causes the error is

textscan( __________ , 'ReturnOnError',false )

It produces a better error message.

请先登录，再进行评论。

请先登录，再回答此问题。

Answer 1

Esha Chakraborty 2022-4-5

0
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/1686674-reading-large-csv-files#answer_934829

Hi Sanchit,

I understand that you are receiving the message - 'Failed to convert character code' when you are attempting to read large CSV files. Possible reason can be that the read buffer is too large and too much data is being read at once. It is suggested to reduce the amount of data being loaded and see if the situation still exists.

Here are a few ways to import large CSV array:

You can try to split the file into smaller sections using any reliable third-party file splitting software, before importing to MATLAB.
You can explore if the Datastore feature suits your use case. A Datastore is an object for reading a single file or a collection of files or data. The Datastore acts as a repository for data that has the same structure and formatting. You can refer to the following documentation page for more details on Datastore here.
You can also explore if MapReduce is an option in your use case. MapReduce is a programming technique for analyzing data sets that do not fit in memory. You can refer to the following documentation page for more details on Mapreduce here.