fix: set inputs as optional

flash/tabular/classification/data/data.py

@@ -71,9 +71,9 @@ class TabularData(DataModule):
     def __init__(
         self,
         train_df: DataFrame,
-        categorical_input: List,
-        numerical_input: List,
         target: str,
+        categorical_input: Optional[List] = None,
+        numerical_input: Optional[List] = None,
         valid_df: Optional[DataFrame] = None,
         test_df: Optional[DataFrame] = None,
         batch_size: int = 2,
@@ -82,6 +82,12 @@ def __init__(
         dfs = [train_df]
         self._test_df = None
 
+        if categorical_input is None and numerical_input is None:
+            raise RuntimeError('Both `categorical_input` and `numerical_input` are None!')
+
+        categorical_input = categorical_input if categorical_input is not None else []
+        numerical_input = numerical_input if numerical_input is not None else []
+
         if valid_df is not None:
             dfs.append(valid_df)
 
@@ -133,8 +139,8 @@ def from_df(
         cls,
         train_df: DataFrame,
         target: str,
-        categorical_input: List,
-        numerical_input: List,
+        categorical_input: Optional[List] = None,
+        numerical_input: Optional[List] = None,
         valid_df: Optional[DataFrame] = None,
         test_df: Optional[DataFrame] = None,
         batch_size: int = 8,
@@ -194,8 +200,8 @@ def from_csv(
         cls,
         train_csv: str,
         target: str,
-        categorical_input: List,
-        numerical_input: List,
+        categorical_input: Optional[List] = None,
+        numerical_input: Optional[List] = None,
         valid_csv: Optional[str] = None,
         test_csv: Optional[str] = None,
         batch_size: int = 8,

tests/tabular/data/test_data.py

@@ -16,6 +16,7 @@
 
 import numpy as np
 import pandas as pd
+import pytest
 
 from flash.tabular import TabularData
 from flash.tabular.classification.data.dataset import _categorize, _normalize
@@ -169,3 +170,11 @@ def test_from_csv(tmpdir):
         assert cat.shape == (1, 1)
         assert num.shape == (1, 2)
         assert target.shape == (1, )
+
+
+def test_empty_inputs():
+    train_df = TEST_DF_1.copy()
+    with pytest.raises(RuntimeError):
+        TabularData.from_df(
+            train_df, categorical_input=None, numerical_input=None, target="label", num_workers=0, batch_size=1
+        )