I need to serialise some Pandas DataFrame data for storing in the JSON. For this purpose I am converting float32
and float64
data types to native Python float
.
- Will there be implications for the data accuracy, or can Python
float
correctly prevent both of these Pandas data types? - Could there be other issues e.g. with
NaN
andInf
? (note: I am already aware that strict JSON does not handleInf
serialisation)