Fix bitsize in parseFloat function

[thevilledev]

parseFloat function calls strconv.ParseFloat with an invalid bitsize parameter. This parameter should either be 32 (for 32-bit precision) or 64 (for 64-bit precision). The result is always of type float64.

Not a user-impacting change, since the internal implementation of strconv.ParseFloat defaults to bitsize of 64 if the given bitsize is anything else than 32.

[eikenb]

Odd error.. and I think you're right that this should have no real impact on users as it would work the same but with the current implementation but is good to fix in case of future changes.