Floating precision encoding - large file

šŸ“Š Plotly Python
1

Iā€™m using Scatter with ndarray.dtype float32. The output file size is not optimal: ā€œyā€: [7.3670160000000005, 7.333488000000001, 7.330439999999999,ā€¦

Are there Plotly JSONification options to avoid creating enormous files when it is not necessary?

2

Please find element of my own answer in case someone else in the community is interested:

def float2str(obj):
    """Float to string."""
    return [format(x, '.5g') for x in obj]


class CustomPlotlyJSONEncoder(PlotlyJSONEncoder):
    """Customized PlotlyJSONEncoder encoder."""

    def default(self, obj):
        """The new custom default encoder."""
        if is_float_dtype(obj):
            return float2str(obj)

        return PlotlyJSONEncoder.default(self, obj)

    sep = (",", ":")
    data_js = dumps(data, cls=CustomPlotlyJSONEncoder, separators=sep)