problems substituting dataread with textscan

Hello everyone

i was given a script which enables me to extract a normalization value from a data file. the data file consist of a header with two columns and a body with three columns of data. the code is shown below

C = dataread('file', 'test.dat', '%s', 'delimiter', '\n') % data into array
string=validatestring('GlobalNormalization=',C) % line with GN into new string
v = sscanf(string,'%*s%f') % puts floating point from the new string into v

It works beautifully but matlab underlines "dataread" and tells me that it will be removed in future releases. Matlab suggests that i instead use tesxtscan. i threfore tried this - the code can be seen below

fid = fopen('test.dat','r');
C = textscan(fid, '%s','delimiter', '\n');  
string=validatestring('GlobalNormalisation=',C);
v = sscanf(string,'%*s%f');
fclose(fid);

this gives me the error message:

"Error using validatestring (line 61) Second argument must be a cell array of allowed strings.

Error in Import_Int (line 9) string=validatestring('GlobalNormalisation=',C);"

I think the reason for this is that dataread simply creates a variable called C wheras textscan creates a 1x1 cell called C. Can anyone tell me how to simply create a variable called C using textscan or suggest some other way of working around the problem?

cheers

Answer 1

[Martin Nors Pedersen]

using

string=validatestring('GlobalNormalisation=',C{1,1});

instead of

string=validatestring('GlobalNormalisation=',C);

solved the problem. i should probably have realized this immediatly. but it got abit confused about the array inside a specific cell position

i am still interested in learning if you can avoid the cell structure and instead just create a variable when you use textscan