Nulls are not properly handled in the first row of the dataset. However, if null exists in a later row, then the chart will render properly.
A current workaround to this problem is to cast null values as strings, but that creates an additional challenge of columns being inferred as "string" type even if they are numeric.
Nulls are not properly handled in the first row of the dataset. However, if null exists in a later row, then the chart will render properly.
A current workaround to this problem is to cast
nullvalues as strings, but that creates an additional challenge of columns being inferred as "string" type even if they are numeric.The source of this issue is here https://github.com/BigFatDog/parcoords-es/blob/master/src/api/detectDimensionTypes.js#L3-L10
I have written this PR with a proposed fix https://github.com/BigFatDog/parcoords-es/pull/117