All the fieldnames are prefixed with 'x'

I'm new to Matlab and I'm sure there's an obvious reason why this isn't working, but I can't figure it out. I've seen this happen in R and SAS sometimes but Google isn't turning up anything.

M = readtable('RRS_with_Chl.csv')

M = 
       x400          x401          x402
    __________    __________    __________
    0.0047333      0.0021455     0.0036116

The csv file that I'm reading from has fieldnames of 400, 401, 402, etc. There are no leading spaces or symbols before them (when I open it in e.g. Notepad++), at least as far as I can tell.

How can i fix this? I would like to extract the fieldnames and use them in a numeric array, but it's hard to do when they have been turned into weird strings.

Edit: I have matlab Version 2015a, so I can't use the strip() function.

Answer 1

[dpb]

table names must be valid Matlab variable names and (just like SAS and R and virtually all other programming languages) that means they must start with a nonnumeric character. readtable has prefixed the letter 'x' to meet that requirement. You can't "fix" that, but you can retrieve the names and obtain the numeric values therefrom simply enough--

N=cellfun(@(s) sscanf(s,'x%d'),M.Properties.VariableNames).';

[Walter Roberson]

Right. In particular, https://www.mathworks.com/help/matlab/ref/matlab.lang.makevalidname.html is used

[dpb]

I'm not sure how readtable does the decision on what the renaming is, if one doesn't have a header row then it uses VarN where N=1,2,...

Whether x is unique for numeric always dunno', a perhaps more robust would be

N=cellfun(@s) sscanf(s(~isletter(s)),'%d',M.Properties.VariableNames).';

although if used an underscore or something would fail on it. Why isn't a builtin in isnumeral dunno', but that'd probably be most robust for OP here.

[Walter Roberson]

Correction, matlab.lang.makeUniqueStrings is used, not matlab.lang.makevalidname