Hi, is there any support in nixtla libraries for target encoding? I believe currently support exists for one hot encoding of categorical features. Please let me know..

Can you clarify a bit more what you mean by target encoding?

Target encoding is technique that we use for dealing with categorical features in time series forecasting/Machine Learning( Just like one hot encoding, label encoding). In this , we replace the categorical feature with the target value mean( or any aggregation).. In some cases, it can be very helpful is what my understanding is.. Further reading: https://www.kaggle.com/code/ryanholbrook/target-encoding#Your-Turn

We don't have explicit encoders in the package, but you can preprocess your features in any way you want before training a model. We also have utilsforecast, which can help you build more features if it's something you want.

ok , cool. thanks