API Guide for Users from Other Data Libraries

Ray Data is a data loading and preprocessing library for ML. It shares certain similarities with other ETL data processing libraries, but also has its own focus. This guide provides API mappings for users who come from those data libraries, so you can quickly map what you may already know to Ray Data APIs.

Note

• This is meant to map APIs that perform comparable but not necessarily identical operations. Select the API reference for exact semantics and usage.

• This list may not be exhaustive: Ray Data isn’t a traditional ETL data processing library, so not all data processing APIs can map to Datasets. In addition, this list focuses on common APIs or APIs that are less obvious to see a connection.

For Pandas Users

Pandas DataFrame vs. Ray Data APIs

Pandas DataFrame API	Ray Data API
df.head()	ds.show(), ds.take(), or ds.take_batch()
df.dtypes	ds.schema()
len(df) or df.shape[0]	ds.count()
df.truncate()	ds.limit()
df.iterrows()	ds.iter_rows()
df.drop()	ds.drop_columns()
df.transform()	ds.map_batches() or ds.map()
df.groupby()	ds.groupby()
df.groupby().apply()	ds.groupby().map_groups()
df.sample()	ds.random_sample()
df.sort_values()	ds.sort()
df.append()	ds.union()
df.aggregate()	ds.aggregate()
df.min()	ds.min()
df.max()	ds.max()
df.sum()	ds.sum()
df.mean()	ds.mean()
df.std()	ds.std()

For PyArrow Users

PyArrow Table vs. Ray Data APIs

PyArrow Table API	Ray Data API
pa.Table.schema	ds.schema()
pa.Table.num_rows	ds.count()
pa.Table.filter()	ds.filter()
pa.Table.drop()	ds.drop_columns()
pa.Table.add_column()	ds.add_column()
pa.Table.groupby()	ds.groupby()
pa.Table.sort_by()	ds.sort()

For PyTorch Dataset & DataLoader Users

For more details, see the Migrating from PyTorch to Ray Data.