I am trying to convert a str column in the DataFrame into Date format of YYYY-MM-DD. How to convert different date format into one format of YYYY-MM-DD?
import polars as pl
s = pl.Series("date",["Sun Jul 8 00:34:60 2001", "12Mar2022", "12/Mar/2022"])
df = s.to_frame().with_columns(pl.col("date").str.to_date("%d%m%Y", strict=False))
shape: (3, 1)
┌──────┐
│ date │
│ --- │
│ date │
╞══════╡
│ null │
│ null │
│ null │
└──────┘
If you have multiple known formats you can use pl.coalesce to combine the results of each parse attempt into a single column.
df = pl.from_repr("""
┌──────────────────────────┐
│ date │
│ --- │
│ str │
╞══════════════════════════╡
│ Sun Jul 8 00:34:60 2001 │
│ 12Mar2022 │
│ 12/Mar/2022 │
└──────────────────────────┘
""")
fmts = "%d/%b/%Y", "%d%b%Y", "%c"
df.with_columns(
pl.coalesce(
pl.col("date").str.to_datetime(fmt, strict=False)
for fmt in fmts
)
)
shape: (3, 1)
┌─────────────────────┐
│ date │
│ --- │
│ datetime[μs] │
╞═════════════════════╡
│ 2001-07-08 00:35:00 │
│ 2022-03-12 00:00:00 │
│ 2022-03-12 00:00:00 │
└─────────────────────┘
