v0.4.7

Fix a deprecated Pandas call.

Fix timestamp truncation in Spark/Parquet

Bug Fix: Fixed an issue with timestamp precision when converting Zeek logs to Parquet format when using Spark.

Corrupt gzip for multi-file reader

Just a patch release for the multi-file reader so that it handles a corrupt gzip file.

Intel Format Support

Small change to support reading in Zeek intel format files (as a dataframe). See: #133

Turbo Snail

Release Notes:

• Various small fixes
• Making pyspark an optional dependency

Sleepy Duckling

• Added a Zeek JSON log to dataframe class (see https://github.com/SuperCowPowers/zat/blob/master/examples/zeek_json_to_pandas.py)

Categorical type ordering corner case fix

There was a small corner case where categorical types would not get properly ordered when doing a fit and the transform. #108

Maintenance Release

Misc code fixes and improvements.

DataFrame Optimization Release

We've optimized how Pandas dataframes are created. The dataframes are created faster and now take up significantly less memory.

Hummingbird Release

• Support for dynamic tailing
• Support for log rotation
• Statistics module
• Lots of new examples, including yara