TypeError: iloc requires integer positions

TypeError: iloc requires integer positions (non-integer key)

$ python -c "import pandas as pd; df = pd.DataFrame({'a':[1,2]}); df.iloc['a']"
Traceback (most recent call last):
  File "<string>", line 1, in <module>
TypeError: cannot do positional indexing with a non-integer key

Why this happens

.iloc is for integer-location based indexing. Using a string (label) or non-integer indexer with .iloc doesn’t make sense because .iloc selects by integer position, not label names.

Fix

Use .loc when you want label-based selection, or use integer positions with .iloc (integers, slices, lists/arrays of integers). Convert labels to integer positions using .get_loc on the index if needed.

Wrong code

import pandas as pd

df = pd.DataFrame({'a': [10, 20]}, index=['x', 'y'])
# Using a label with iloc -> TypeError

val = df.iloc['x']

Fixed code

import pandas as pd

df = pd.DataFrame({'a': [10, 20]}, index=['x', 'y'])
# Use label-based selection with .loc
val = df.loc['x']

# Or use integer positions with .iloc
val0 = df.iloc[0]

# If you need to translate a label to integer position
pos = df.index.get_loc('x')
val_using_pos = df.iloc[pos]