Is there a way to configure the %d in fprintf function to

John Michell answered . 2025-08-17 23:14:38

It looks like this might have something to do with flintmax, which would mean there would be no way to change this. This value is 2^53 for doubles, which works out to 9007199254740992 (so just under 17 digits). If I try writing more digits, my copy of Matlab starts rounding. I also observe the conversion to exponential notation at 20 digits. Hidden in the large doc for fprintf, you can find this:

 

 

If you specify a conversion that does not fit the data, such as a text conversion for a numeric value, 
MATLAB overrides the specified conversion, and uses %e.
Example: '%s' converts pi to 3.141593e+00.

%replace this:
%fprintf(not,'%21d', step(k,l));

%before the loop:
FormatSpec={'%21d','%21e'};
NumberOfDigits=@(x) 1+floor(log10(x));

%in the loop
val=step(k,l)
fprintf(not,FormatSpec{1+(val>flintmax)}, val);
%or:
%fprintf(not,FormatSpec{1+(NumberOfDigits(val)>19)}, val);

 

Not satisfied with the answer ?? ASK NOW