Merge pull request #6 from amazon-science/adivekar

[bears] Migrated data processing pipelines from fmcore to bears
amazon-science · Jan 25, 2025 · d159d89 · d159d89
2 parents 0124135 + 63d4fb9
commit d159d89
Show file tree

Hide file tree

Showing 26 changed files with 2,423 additions and 1 deletion.
diff --git a/.gitignore b/.gitignore
@@ -19,4 +19,4 @@ __pycache__/
 *.swp
 
 .vscode/settings.json
-Test-bears.ipynb
+*.ipynb
diff --git a/pyproject.toml b/pyproject.toml
@@ -34,6 +34,7 @@ dependencies = [
     "tqdm",
     "boto3",
     "cloudpickle>=3.0.0",
+    "scikit-learn",
 ]
 
 [project.optional-dependencies]

diff --git a/src/bears/__init__.py b/src/bears/__init__.py
@@ -10,6 +10,7 @@
 from bears.core.frame.ScalableSeries import ScalableSeries, ScalableSeriesRawType,ScalableSeriesOrRaw
 from bears.reader import Reader 
 from bears.writer import Writer
+from bears.processor import DataProcessor, DataPipeline
 
 to_sdf = ScalableDataFrame.of
 to_ss = ScalableSeries.of
-Original file line number
+Diff line change
@@ Expand Up / @@ -19,4 +19,4 @@ __pycache__/ @@
     *.swp
     .vscode/settings.json
-    Test-bears.ipynb
+    *.ipynb