When I use VectorAssembler and xgboost4j-spark, I get “you can only specify missing value as 0.0 (the currently set value NaN) when you have SparseVector or Empty vector as your feature format”. Dose ‘missing=0.0’ mean treating 0.0 as missing value. However, I hope that 0 in SparseVector be treated as 0 not missing value.
By the way, I have ensured that there are no missing values in the original data.
My xgboost4j-spark version is 0.90.
Can anyone help me? Thanks.
How to understand missing=0.0 parameter?
I understand. ‘missing=0.0’ mean treating 0.0 as missing value, and it will not affect training results. Missing value also will be be classified like normal value.