pandas删除含NaN行报错：TypeError无效切片键问题求助

Ah, I see the issue here! Your error comes from using numpy-style 2D slicing (data[:59,:]) with pandas' drop() method, which doesn't accept that kind of key. The drop() function expects row labels (indices) or a boolean mask, not a slice like that. Let's fix this with a few straightforward solutions:

Solution 1: Drop rows by index range #

Since you know exactly which rows (0 to 58) need removal, you can pass the range of indices directly to drop():

# Drop rows 0 to 58 (inclusive)
data.drop(index=range(59), inplace=True)
print(data)

If you prefer not to modify the original DataFrame in-place:

data = data.drop(index=range(59))

Solution 2: Slice the DataFrame directly (simpler!) #

A more concise way to keep only rows from index 59 onwards is to use pandas' row slicing:

# Keep all rows starting from index 59
data = data[59:]
print(data)

This works because data[start:] returns every row from the start index to the end of the DataFrame, effectively removing the first 59 rows.

Solution 3: Drop rows with NaN in specific columns (more robust) #

If you want to ensure any row with NaN in Real Lower Band or Real Upper Band is removed (not just the first 59 rows), use dropna() with the subset parameter:

# Remove rows where either of these columns has NaN
data.dropna(subset=['Real Lower Band', 'Real Upper Band'], inplace=True)
print(data)

This is a better approach if there might be other rows beyond index 58 that also contain NaN values in those columns.

内容的提问来源于stack exchange，提问作者Samar Pratap Singh