zingg
zingg copied to clipboard
zinggAI
Scalable identity resolution, entity resolution, data mastering and deduplication using ML
currently it just gives an npe and it is difficult for the user to make out whats going wrong.
many helper functions in the notebooks can be shifted to Zingg package so that notebooks are neater
Column width should be auto deduced. Sometimes the valeus are overlapping  Color scheming can be better We can show pos and neg separately, right now they are appearing as...
Currently we get NPEs at many places which appear as stack traces and no helpful messages are printed in the log. We should define better exception handling and provide messages...
Let us build automated tests for loads of febrl120k, febrl500k, ncVoter 5m and febrl5m and send a weekly report so we can track perf degradations or lifts.
1. Bring Enterprise Greedy Optimisations to Oepn Source 2. Investigate dupeN setting and why it is not being changed from parent to child 3. Cache info in isFunctionUsed 4. If...
Sometimes Zingg jobs are slow or fail due to a poorly learnt blocking tree. This can happen due to a variety of reasons. For example when a user adds sgnificantly...
the model has z_source whereas the code now expects z_zsource