I can only see append and overwrite modes in writing to delta table. but I cant find anything to upsert to existing delta table without having to overwrite the table completely.

We don’t support update yet, but it has been started. We do support overwriting individual partitions, though.

Thanks Will for the response. When you say individual partition overwrite, do you mean this -- write_deltalake(table, df2, partition_filters=[('y', '=', 'b')], mode="overwrite") ?

Yes, that’s it

example which was given in the documentation, is somehow not working as expected.

from deltalake.writer import write_deltalake
>>> df = pd.DataFrame({'x': [1, 2, 3], 'y': ['a', 'a', 'b']})
>>> write_deltalake('path/to/table', df, partition_by=['y'])

>>> table = DeltaTable('path/to/table')
>>> df2 = pd.DataFrame({'x': [100], 'y': ['b']})
>>> write_deltalake(table, df2, partition_filters=[('y', '=', 'b')], mode="overwrite")

>>> table.to_pandas()
     x  y
0    1  a
1    2  a
2  100  b

in this example as it should overwrite the file of partition y, but this is not updating the partition file .

The code you are showing seems to have the correct result. Are you getting different data at the end?