New Column.__repr__() for an nice overview of the column

Author: FrancescAlted

examples/ctable/indexing.py

@@ -41,10 +41,12 @@ def load_rows(table: blosc2.CTable, nrows: int = 240) -> None:
 packed = None
 
 try:
+    print("Creating a CTable with mixed dtypes...")
     pt = blosc2.CTable(Measurement, urlpath=str(table_path), mode="w")
     load_rows(pt)
 
     # Create a couple of indexes on columns with different dtypes.
+    print("\nCreating indexes...")
     idx_sensor = pt.create_index("sensor_id", kind=blosc2.IndexKind.FULL)
     idx_active = pt.create_index("active")
     print("Indexes created:", pt.indexes)
@@ -54,8 +56,8 @@ def load_rows(table: blosc2.CTable, nrows: int = 240) -> None:
     # Queries can combine indexed and non-indexed predicates.
     recent_active = pt.where((pt["sensor_id"] >= 180) & pt["active"] & (pt["region"] == "north"))
     print("\nLive rows with sensor_id >= 180, active=True, region='north':", len(recent_active))
-    print("sensor_ids:", recent_active["sensor_id"].to_numpy().tolist())
-    print("statuses:", recent_active["status"].to_numpy().tolist())
+    print("sensor_ids:", recent_active["sensor_id"])
+    print("statuses:", recent_active["status"].to_numpy())
 
     # Close the table, pack the TreeStore into a single .b2z file, and reopen it.
     del pt
@@ -79,8 +81,8 @@ def load_rows(table: blosc2.CTable, nrows: int = 240) -> None:
     # Query directly against the .b2z bundle; no unpack step is needed.
     warm_active = packed.where(packed["active"] & (packed["status"] == "warm") & (packed["sensor_id"] > 100))
     print("\nRows from .b2z with active=True, status='warm', sensor_id > 100:", len(warm_active))
-    print("sensor_ids:", warm_active["sensor_id"].to_numpy().tolist())
-    print("regions:", warm_active["region"].to_numpy().tolist())
+    print("sensor_ids:", warm_active["sensor_id"])
+    print("regions:", warm_active["region"].to_numpy())
 
     print("\nThe packed file is kept on disk.")
     print(f"Inspect it later with: f = blosc2.open({bundle_path.name!r}, mode='r')")

src/blosc2/ctable.py

@@ -12,6 +12,7 @@
 
 import contextlib
 import dataclasses
+import itertools
 import os
 import pprint
 import shutil
@@ -475,6 +476,8 @@ def __getitem__(self, col_name: str):
 
 
 class Column:
+    _REPR_PREVIEW_ITEMS = 8
+
     def __init__(self, table: CTable, col_name: str, mask=None):
         self._table = table
         self._col_name = col_name
@@ -597,6 +600,21 @@ def __iter__(self):
             data_chunk = self._raw_col[chunk_start : chunk_start + actual_size]
             yield from data_chunk[mask_chunk]
 
+    def __repr__(self) -> str:
+        preview_items = []
+        for value in itertools.islice(self, self._REPR_PREVIEW_ITEMS + 1):
+            if isinstance(value, np.generic):
+                value = value.item()
+            preview_items.append(repr(value))
+
+        truncated = len(preview_items) > self._REPR_PREVIEW_ITEMS
+        if truncated:
+            preview_items = preview_items[: self._REPR_PREVIEW_ITEMS]
+            preview_items.append("...")
+
+        preview = ", ".join(preview_items)
+        return f"Column({self._col_name!r}, dtype={self.dtype}, len={len(self)}, values=[{preview}])"
+
     def __len__(self):
         return blosc2.count_nonzero(self._valid_rows)
 

tests/ctable/test_column.py

@@ -675,5 +675,15 @@ def test_repr_is_single_line():
     assert "\n" not in repr(t)
 
 
+def test_column_repr_shows_preview_values():
+    t = CTable(Row, new_data=DATA20)
+    r = repr(t["id"][:])
+    assert "Column('id'" in r
+    assert "dtype=int64" in r
+    assert "len=20" in r
+    assert "values=[0, 1, 2" in r
+    assert "..." in r
+
+
 if __name__ == "__main__":
     pytest.main(["-v", __file__])