Selecting & Filtering Data

.loc and .iloc are how you pull specific rows and columns out of a DataFrame — one works by label, the other by position, and mixing them up is the most common early source of confusion in pandas.

Selecting a column

>>> a column

import pandas as pd

df = pd.DataFrame({
    "name": ["Maya", "Sam", "Priya", "Leo"],
    "score": [88, 92, 79, 95]
})
print(df["score"])

Output

0    88
1    92
2    79
3    95
Name: score, dtype: int64

df["score"] pulls out a single column as a Series. df.score works too when the column name is a valid Python identifier, but df["score"] is the safer habit — it also works for column names with spaces or that collide with DataFrame method names.

.loc: select by label

.loc[row_labels, column_labels] selects using the actual index and column names:

>>> label-based selection

print(df.loc[1, "name"])
print(df.loc[0:2, ["name", "score"]])

Output

Sam
    name  score
0   Maya     88
1    Sam     92
2  Priya     79

.iloc: select by position

.iloc[row_positions, column_positions] ignores labels entirely and counts from 0, exactly like list indexing:

>>> position-based selection

print(df.iloc[1, 0])
print(df.iloc[0:2, :])

Output

Sam
   name  score
0  Maya     88
1   Sam     92

The gotcha: when the index is the default 0, 1, 2… .loc and .iloc often return the same thing, which hides the difference — until you filter, sort, or set a custom index, at which point row label 5 and row position 5 can point at completely different rows. Prefer .iloc when you mean "the Nth row" and .loc when you mean "the row labeled X."

Boolean filtering

Passing a True/False Series back into df[...] keeps only the rows where it's True — this is how you filter a DataFrame with a condition:

>>> filtering

print(df[df["score"] >= 90])

Output

  name  score
1  Sam     92
3  Leo     95

Combining conditions uses & and |, not Python's and/or — and each condition needs its own parentheses:

>>> combined condition

print(df[(df["score"] >= 90) & (df["name"] != "Sam")])

Output

  name  score
3  Leo     95

PreviousReading Data NextModifying Data