TEXTSCAN problem using TREATASEMPTY

D = textscan(fid,'%q%f%f%f%f%f%f%f%f%f%f%f%f%f%f','Delimiter',',','HeaderLines',4,'TreatAsEmpty',{'"NAN"','"INF"','"-INF"','"1.#INF"','"-1.#IND"','"#NAME?"'});

This fails.

D = textscan(fid,'%q%f%f%f%f%f%f%f%f%f%f%f%f%f%f','Delimiter',',','HeaderLines',4,'TreatAsEmpty',{'"NAN"'});

This works.

The file in question looks much like the following line.

"2011-04-19 11:55:30",521882,1,25.5,347.068,3.765646,1.54899,18.43283,204.205,125.9697,10,651.2458,9.310149,"NAN","NAN"

The last two columns are all "NAN" values. When using the first code example Matlab loads all data except for the very last value. If the very last value is replaced by a number it loads everything.

What is going on with TEXTSCAN?

[Jeremy Hughes]

Try adding

'ReturnOnError',false

to get a better diagnostic of why textscan stops there.