What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
iTechGuides is reader-supported. When you buy through links on our site, we may earn an affiliate commission. As an Amazon Associate I earn from qualifying purchases. Learn more
Choose a pandas selector by asking two questions: are you identifying rows and columns by labels or by integer positions, and do you need one value or a larger selection? Use [] for a single column or a simple row filter, .loc for label-based selections, .iloc for zero-based positional selections, and .at or .iat for a single scalar.
How do I select a subset of a DataFrame?
Start with this small DataFrame. Its index labels are strings, so the distinction between a label and a row position is visible:
import pandas as pd
df = pd.DataFrame(
{
"name": ["Ada", "Linus", "Grace"],
"age": [36, 34, 40],
"city": ["London", "Helsinki", "New York"],
},
index=["row_a", "row_b", "row_c"],
)For example, row_a is the label of the first row, while 0 is its zero-based position. The core distinction is:
Quick wins for a faster PC:
Scan for outdated or missing drivers - takes under a minuteDriver Scan →Clear out junk files and repair common Windows errorsFree Scan →Fix the driver behind crashes, sound loss and screen glitchesFind Drivers →.locuses index and column labels..ilocuses integer positions, counted from zero..atand.iatuse the same label-versus-position distinction for one value.
Use [] for a column or a straightforward row filter
Select a column
Pass the column label in brackets to get a Series:
df["name"]This returns the name values for all rows. To select more than one named column, pass a list of labels; the result is a DataFrame:
#1 Best Overall
df[["name", "age"]]Filter rows with a condition
A boolean condition inside brackets keeps rows where the condition is true:
df[df["age"] > 35]The condition produces a Boolean Series aligned with the DataFrame’s rows. You can also use an explicit label-based selection when you want to state the row condition and chosen columns together:
df.loc[df["age"] > 35, "name"]This returns the names of people older than 35. The pandas getting-started tutorial demonstrates this pattern with its Titanic data, selecting names where Age is greater than 35 (pandas getting-started tutorial).
Do these 3 things before closing this tab:
1Clear out junk files and repair common Windows errors2Fix the driver behind crashes, sound loss and screen glitches3Repair Windows errors before they cause bigger problemsUse .loc when selectors are labels
The general form is df.loc[row_selector, column_selector]. The selector before the comma addresses rows; the selector after it addresses columns. Use : when you want all rows or all columns on one axis.
# Rows by index labels, and columns by names
df.loc["row_a", "name"]
# All rows, selected columns
df.loc[:, ["name", "age"]]
# A range of row labels and two columns
df.loc["row_a":"row_c", ["name", "age"]]Unlike ordinary Python slicing, a .loc label slice includes its stop label when that label is present. Thus "row_c" is included in the last example. A number passed to .loc is still interpreted as an index label, not as a row number. If a requested label does not exist, .loc raises KeyError.
Use .iloc when selectors are integer positions
The form is df.iloc[row_position, column_position]. Positions start at zero: position 0 is the first row or column. For example:
# First row, second column
df.iloc[0, 1]
# First three rows, second and third columns
df.iloc[0:3, [1, 2]]In the second example, the row slice includes positions 0, 1, and 2, while the selected column positions are 1 and 2. Positional slices follow Python and NumPy rules: the start is included and the stop is excluded. A scalar or list request for a position outside the axis raises IndexError; slice endpoints may extend out of bounds without that error.
Use .at and .iat to get or set one value
When the desired result is one scalar rather than a Series or DataFrame, use the scalar accessor that matches your selector:
| Accessor | Selector meaning | Example |
|---|---|---|
.at |
Row and column labels | df.at["row_a", "age"] |
.iat |
Zero-based row and column positions | df.iat[0, 1] |
Both examples retrieve Ada’s age from this DataFrame. These accessors can also assign a value, for example df.at["row_a", "age"] = 37. Assignment through .at to a missing indexer can enlarge the object in place; take care when an apparent typo could create a new entry. The pandas API reference documents .iat as the integer-position scalar accessor (DataFrame.iat API reference).
Which selector should you use?
| Need | Use | What to remember |
|---|---|---|
| One named column | df["column"] |
Returns a Series. |
| Rows matching a simple condition | df[condition] |
Boolean filtering keeps matching rows. |
| A selection by row or column labels | df.loc[rows, columns] |
Label slices include their stop label when present. |
| A selection by row or column positions | df.iloc[rows, columns] |
Positions start at zero; slice stops are excluded. |
| One value by labels | df.at[row_label, column_label] |
Scalar label lookup. |
| One value by positions | df.iat[row_position, column_position] |
Scalar positional lookup. |
The [] syntax is convenient for columns and uncomplicated filtering. For production code, the pandas user guide recommends the explicit access methods .at, .iat, .loc, and .iloc; it does not supply a benchmark or a numeric performance advantage for these methods (pandas indexing and selecting data guide). For version-specific details, consult the documentation matching the pandas version installed in your environment; the guide and API references may describe different releases as they are updated.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
