Does converting a DeltaTable into a Pandas DataFrame downloads the whole table?

[this]

I have the following Python code to connect to a Delta table and convert that into a Pandas DataFrame.

from deltalake import DeltaTable

dt = DeltaTable(path, storage_options=storage_options)
df = dt.to_pandas()

As I'm executing the above code remotely, will it download the whole table to create the DataFrame?

[riziles]

I thought pandas was in-memory only with no concept of lazy data? Doesn't that imply that it would have to?

[thomas-chauvet]

Yes, it will. It uses pyarrow underneath to read everything if you do not pass any filter. You can pass filters or partitions for even better efficiency.