Remove precision from dump float (#245)

Precision in a python format string, without a type specifier, formats
the content as a _string_, where the precision indicates the maximum
field size. This results in numbers with a large integer part, but a
small decimal part, being truncated unnecessarily.
1 file changed