Kahan summation: Difference between revisions
→{{header|Julia}}
Line 1,080:
=={{header|Julia}}==
Julia can use its BigFloat data type to avoid floating point epsilon errors, as shown below.
<lang julia>epsilon() = begin eps = 1.0; while 1.0 + eps != 1.0
function kahansum(arr)
Line 1,109:
BigFloat sum is 1.0
</pre>
=={{header|Kotlin}}==
| |||