Uniform report formatting
Prepare tabular numeric output for reports by applying one consistent decimal precision.
Pandas · Module
df.round
Rounds numeric DataFrame columns to a chosen precision. Use an integer for uniform precision or a dict/Series for per-column precision.
How it works
Call round with an integer to apply one precision to every numeric column, or pass a dictionary or Series to choose precision per column. The operation returns a DataFrame; columns omitted from a mapping remain unchanged, and non-numeric columns are not rounded.
Syntax
DataFrame.round(decimals=0, *args, **kwargs)Parameters
| Name | Type | Description |
|---|---|---|
decimals | int, dict, or Series | Number of decimal places for each column. An integer applies uniformly; a dict or Series specifies per-column precision. |
**args, **kwargs | optional | Additional keyword arguments accepted for compatibility with NumPy; they have no effect. |
Returns
DataFrame with the affected columns rounded to the specified number of decimal places.
Key points
- DataFrame.round produces a DataFrame with numeric values rounded to requested decimal places.
- Passing one integer applies uniform precision across numeric columns.
- A dictionary or Series enables per-column precision.
- Dictionary-like inputs identify target columns through their keys.
- Series inputs identify target columns through their index.
- Unspecified columns retain their existing values.
Common mistakes
- Halfway cases use ties-to-even rather than always rounding away from zero.
- Misspelled or extra column names in a mapping do not raise an error; they are ignored.
- Do not rely on extra keyword arguments to alter the rounding behavior.
Example: Round selected columns
import pandas as pd
df = pd.DataFrame({
"price": [1.234, 5.678],
"quantity": [2.345, 6.789],
"label": ["a", "b"],
})
rounded = df.round({"price": 2, "quantity": 1})
print(rounded)Rounds selected numeric columns independently while leaving the text column unchanged.
When to use it
Column-specific precision
Round measurements, prices, or ratios with different precision requirements per column.
Non-destructive transformation
Create a presentation-ready copy while preserving the original unrounded DataFrame.
Interview questions
QHow can different columns receive different precision?
An integer applies the same number of decimal places to every numeric column. A dict maps column names to individual precisions, while a Series uses its index as the column-name mapping.
Follow-up: What happens to columns omitted from a dict? They remain unchanged, so a mapping can target only the columns requiring a different precision.
QWhat does DataFrame.round return?
The method returns a DataFrame containing the affected columns rounded to the requested precision. It does not modify the original DataFrame in place.
Follow-up: How would you preserve the original while using the result? Assign the returned DataFrame to another variable, as in rounded = df.round(2).
QWhat forms can the decimals argument take?
The decimals argument accepts an integer, dictionary, or Series. Dictionary keys must be column names; for a Series, the relevant column names must be in its index.
Follow-up: What happens if a mapping includes a name absent from the DataFrame? That entry is ignored.
Related
DataFrame.absDataFrame.clipDataFrame.astype