How to stop matlab from truncating long decimal numbers

Is there a way for a long string of decimal numbers like 0.92834387282837738 to not be truncated by matlab?? I was using it for a code (arithmetic coding) that requires precision and matlab truncating the numbers ruins everything.

Answer 1

[madhan ravi]

The decimals are not truncated it’s just the appearance in command window:

doc format
doc sprintf

[Rik]

There is also actual truncation if the values don't have a perfect binary representation. Float rounding is something to be aware of if you need a lot of precision.

[Walter Roberson]

By my count the sample has more than 16 digits. There is no way to prevent truncation to 53 bits (15 to 16 digits).

If the user has the symbolic toolbox then

sym('0.92834387282837738', 'd')

would not be truncated. Longer strings could require setting the digits() first.

[Giuseppe Degan Di Dieco]

Hello Guys,

thanks for setting up this thread, it really helped me.

Indeed, the command

sprintf

is very useful to create string that meets the user's needs.

Best.

[Steven Lord]

No, sprintf won't work if the number is truncated by being stored as a double precision number before it gets to sprintf.

sprintf('%0.100f', 0.0123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789)

ans = '0.0123456789012345684308780136007044347934424877166748046875000000000000000000000000000000000000000000'

If as Walter Roberson suggested you'd converted this directly from text to a sym object:

s = sym('0.0123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789');
vpa(s, 100)

ans = 

0.0123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789