How to split an array cell with multiple text line with carriage return

I have ingested an excel file that has one cell with multiple strings with each separated by carriage return...ie...

abc(carriage return symbol)efg(carriage return symbol)hij(cr symbol)...ect in on cell array. As a result, when I say array(i) has a value of abc(carriage return symbol)efg(carriage return symbol)hij(cr symbol). I want to have array(1) have values of abc array(2) of efg.......ect. Separate each text by the Carriage return symbol at the end and put it in new cell.

Answer 1

[dpb]

We can't test without actual data sample to be sure what is actually embedded in the string, but if it's using standard whitespace, then simply

array=split(array);

should do the trick.

If it doesn't, attach the actual array as a .mat file or the Excel file itself.

Depending upon the content of that file and how you read it, you may be able to avoid the need by using proper syntax/options with which to read the file to begin with.

[Nischal Amin]

That didn't work. The array I have is size of 1x29 of string array...

x = [abc(cr)def(cr)ghi, jkl(cr), mno(cr)pqr(cr)stu(cr), ..... ect to col#29]

I want above x to be like...

x = [abc, def, ghi, jkl, mno, pqr, stu .... all the way to col#29].

x=split(x) didn't work

I exported an excel file into Matlab. Some cell have multiple strings line break/cr.

[dpb]

We can't debug what we can't see/touch.

As noted, attach either the Excel file or at least the .mat file containing the array values.

x = ['abc' newline 'def' newline 'ghi'];
whos x

  Name      Size            Bytes  Class    Attributes

  x         1x11               22  char               

x=split(x)

x = 3×1 cell array
    {'abc'}
    {'def'}
    {'ghi'}

certainly works as advertised -- ergo, there must be something else in the actual string than a recognized whitespace character. But, we can't tell what without the data itself.

Show all your code to read the file and attach the file if you want help...

[Nischal Amin]

I have attached two files. Run Matlab script and pick the excel file. I would like X to be...

x = [abc, def, ghi, jkl, mno, pqr, stu, vwx, yz, ABC]. Size of x = 1x10.

It's showing...

x = {'mno'} {'pqr'} {'stu'}

{'vwx'} {'yz' } {'abc'}

X size of 2 x 3

[dpb]

The problem there is in the format of the input file -- you've got stuff scattered around all over creation in multiple cells and expecting it to be smooshed all back together again.

Why would you do that? If all the strings are supposed to be in one column, then putting them there in the input file into a single column would be the logical thing to do.

It can be forced to do so, but it's ugly...

My release (R2020b) won't read string data with readmatrix so I used readcell instead, but one way to beat it into the desired form is..

data=readcell('try.xlsx','NumHeaderLines',1);
data=data(cellfun(@(c)all(~ismissing(c)),data,'UniformOutput',1));
data=split(join(data,newline));

But, I'd still strongly recommend to not stuff your data into such an unstructured format. Obviously, this is just an example of some other "real" file structure, and maybe it's not possible to avoid such a mess, but do so if possibly can.

split unfortunately, doesn't have a way to handle the different number of delimiters in the various cells and so one gets into an infinite loop of cell arrays by any application of it inside cellfun or arrayfun because the 'uniformoutput',0 flag has to be set and so one just gets back the same cell array form started with.

So, the "trick" is to string all the cells together, also joining them with a newline and split() the resulting single long string...

It is that single string which is the format you showed in the original post that led to the provided answer -- but, that showed that that wasn't actually the data that had in that form -- yet.