Skip to content

pandas/strftime-date-part

Rule strftime-date-part
Pack pandas
Severity warning
Fix unsafe, with xpd check --fix --unsafe-fixes
Library versions every version

Formatting dates as text and parsing a number back out was 33 to 198 times slower than reading the part. Read it through .dt, as in df["d"].dt.year.

year = df["time"].dt.strftime("%Y").astype(int)
weekday = df["time"].dt.strftime("%A")
year = df["time"].dt.year
weekday = df["time"].dt.day_name()

The rule reports df["d"].dt.strftime("%Y").astype(int), the same with %m, %H and %j, and dt.strftime("%A") and "%B", which were 33 to 198 times slower than the date part on 1,000,000 rows.

xpd check --fix --unsafe-fixes applies the fixes of this rule. Check the result, because an unsafe fix can change what the code does.

The unsafe fix writes df["d"].dt.year, df["d"].dt.day_name().

The fixes are unsafe, because a date part is int32 where astype(int) gave int64, and day_name is English where strftime follows the locale.

xpd prints this message with the issue. A name that starts with $ stands for code from your file.

This formats every date as text and parses the text back into a number. Read the part with $S.dt.year.

Browse the other rules of the pandas pack.