You do not need to do that. There some other approaches you can take:
1) readtable() the file as-is. In R2018b at least, that will return a table with 3 variables named Var1, Var2, Var3 -- the original header will be gone . If you need to, you can assign the names to the Properties.VariableNames property of the table
2) Similar to the above, but you also
T = readtable('test.csv');
fid = fopen('test.csv', 'rt');
S = fgetl(fid);
fclose(fid);
varnames = regexp(S, ',', 'split);
varnames( cellfun(@isempty, varnames) ) = []; %remove any trailing empty fields
T.Properties.VariableNames = matlab.lang.makeUniqueStrings(matlab.lang.makeValidName(varnames));
3) Use textscan() to read the file, either with or without bothering to extract the variable names
4) If you do not need the variable names,
csvread('test.csv', 1, 0)
The 1 tells it to skip 1 row; the 0 tells it to skip zero columns.
