BUG: KeyMap has a failure with pandas 3.0.3

[VisLab]

Issue: KeyMap._remap() ValueError on pandas 3.0.3

KeyMap._remap() fails with ValueError: Length of values (2) does not match length of index (0) on pandas 3.0.3

---

Exact Reproduction Case

Input Data

import pandas as pd

# Test DataFrame - exactly as used in test_numeric_keys_cascade
df = pd.DataFrame({
    'test': [1, 2, 'n/a', 3, 4, 5],
    'response_accuracy': ['correct', 'correct', 'correct', 'n/a', 'correct', 'correct'],
    'response_hand': ['right', 'right', 'right', 'n/a', 'left', 'left']
})

print("Input DataFrame:")
print(df)
print("\nDataFrame dtypes:")
print(df.dtypes)
print("\nDataFrame shape:", df.shape)

Output:

Input DataFrame:
  test response_accuracy response_hand
0    1            correct         right
1    2            correct         right
2  n/a            correct         right
3    3                n/a           n/a
4    4            correct          left
5    5            correct          left

DataFrame dtypes:
test                  object
response_accuracy     object
response_hand         object
dtype: object

DataFrame shape: (6, 3)

Operation Configuration

from hed.tools.analysis.key_map import KeyMap

# Remap configuration - cascading numeric keys
# Maps: 1→"correct_left", 2→"correct_right" for test column
map_list = [
    [1, 1, "correct_left"],      # When test=1 AND response_accuracy=1, create "correct_left"
    [2, 2, "correct_right"]      # When test=2 AND response_accuracy=2, create "correct_right"
]

key_map = KeyMap(
    source_columns=['test'],
    map_list=map_list
)

# This is where it fails
df_remapped, missing = key_map.remap(df)

---

Exact Error

On pandas 3.0.3 (FAILS):

ValueError: Length of values (2) does not match length of index (0)