Open frozenkp opened 7 years ago
For anybody else that comes across this issue, it's possible I think (?) with the https://godoc.org/github.com/sjwhitworth/golearn/base#Filter and the https://godoc.org/github.com/sjwhitworth/golearn/base#LazilyFilteredInstances type. Basically, you create a function which takes a []byte
, unpacks it to float64
, transforms it, and then returns a float64
marshalled back as a []byte
. In terms of standard scaling etc, we don't have any utility functions which can do this at the moment.
Is there any way to normalize data with golearn? If so, how can I do it?