org.sonar.l10n.py.rules.python.S6740.html Maven / Gradle / Ivy
This rule raises an error when the dtype parameter is not provided when using pandas.read_csv or
pandas.read_table.
Why is this an issue?
The pandas library provides an easy way to load data from documents hosted locally or remotely, for example with the pandas.read_csv
or pandas.read_table functions:
import pandas as pd
df = pd.read_csv("my_file.csv")
Pandas will infer the type of each columns of the CSV file and specify the datatype accordingly, making this code perfectly valid. However this
snippet of code does not convey the proper intent of the user, and can raise questions such as:
- What information can I access in
df?
- What are the names of the columns available in
df?
These questions arise as there are no descriptions of what kind of data is loaded into the data frame, making the code less understandable and
harder to maintain.
A straightforward way to fix these issues is by providing the schema of the data through the usage of the dtype parameter.
How to fix it
To fix this issue provide the dtype parameter to the read_csv or read_table function.
Code examples
Noncompliant code example
import pandas as pd
def foo():
return pd.read_csv("my_file.csv") # Noncompliant: it is unclear which type of data the data frame holds.
Compliant solution
import pandas as pd
def foo():
return pd.read_csv(
"my_file.csv",
dtype={'name': 'str', 'age': 'int'}) # Compliant
Resources
Documentation
- Pandas Documentation - pandas.read_csv
- Pandas Documentation - pandas.read_table
- Pandas Documentation - dtypes