Skip to content

Perform joins on nearest keys

Source code

Description

This is similar to a left-join except that we match on nearest key rather than equal keys.

Usage

<LazyFrame>$join_asof(
  other,
  ...,
  left_on = NULL,
  right_on = NULL,
  on = NULL,
  by_left = NULL,
  by_right = NULL,
  by = NULL,
  strategy = c("backward", "forward", "nearest"),
  suffix = "_right",
  tolerance = NULL,
  allow_parallel = TRUE,
  force_parallel = FALSE
)

Arguments

other LazyFrame
… Not used, blocks use of further positional arguments
left_on, right_on Same as on but only for the left or the right DataFrame. They must have the same length.
on Either a vector of column names or a list of expressions and/or strings. Use left_on and right_on if the column names to match on are different between the two DataFrames.
by_left, by_right Same as by but only for the left or the right table. They must have the same length.
by Join on these columns before performing asof join. Either a vector of column names or a list of expressions and/or strings. Use left_by and right_by if the column names to match on are different between the two tables.
strategy Strategy for where to find match:
  • "backward" (default): search for the last row in the right table whose on key is less than or equal to the left key.
  • "forward": search for the first row in the right table whose on key is greater than or equal to the left key.
  • "nearest": search for the last row in the right table whose value is nearest to the left key. String keys are not currently supported for a nearest search.
suffix Suffix to add to duplicated column names.
tolerance Numeric tolerance. By setting this the join will only be done if the near keys are within this distance. If an asof join is done on columns of dtype "Date", "Datetime", "Duration" or "Time", use the Polars duration string language. About the language, see the Polars duration string language section for details. There may be a circumstance where R types are not sufficient to express a numeric tolerance. In that case, you can use the expression syntax like tolerance = pl$lit(42)$cast(pl$Uint64)
allow_parallel Allow the physical plan to optionally evaluate the computation of both DataFrames up to the join in parallel.
force_parallel Force the physical plan to evaluate the computation of both DataFrames up to the join in parallel.

Details

Both tables (DataFrames or LazyFrames) must be sorted by the asof_join key.

Polars duration string language

Polars duration string language is a simple representation of durations. It is used in many Polars functions that accept durations.

It has the following format:

  • 1ns (1 nanosecond)
  • 1us (1 microsecond)
  • 1ms (1 millisecond)
  • 1s (1 second)
  • 1m (1 minute)
  • 1h (1 hour)
  • 1d (1 calendar day)
  • 1w (1 calendar week)
  • 1mo (1 calendar month)
  • 1q (1 calendar quarter)
  • 1y (1 calendar year)

Or combine them: “3d12h4m25s” # 3 days, 12 hours, 4 minutes, and 25 seconds

By "calendar day", we mean the corresponding time on the next day (which may not be 24 hours, due to daylight savings). Similarly for "calendar week", "calendar month", "calendar quarter", and "calendar year".

Examples

library(polars)

#
# create two LazyFrame to join asof
gdp = pl$LazyFrame(
  date = as.Date(c("2015-1-1", "2016-1-1", "2017-5-1", "2018-1-1", "2019-1-1")),
  gdp = c(4321, 4164, 4411, 4566, 4696),
  group = c("b", "a", "a", "b", "b")
)

pop = pl$LazyFrame(
  date = as.Date(c("2016-5-12", "2017-5-12", "2018-5-12", "2019-5-12")),
  population = c(82.19, 82.66, 83.12, 83.52),
  group = c("b", "b", "a", "a")
)

# optional make sure tables are already sorted with "on" join-key
gdp = gdp$sort("date")
pop = pop$sort("date")


# Left-join_asof LazyFrame pop with gdp on "date"
# Look backward in gdp to find closest matching date
pop$join_asof(gdp, on = "date", strategy = "backward")$collect()

#> shape: (4, 5)
#> ┌────────────┬────────────┬───────┬────────┬─────────────┐
#> │ date       ┆ population ┆ group ┆ gdp    ┆ group_right │
#> │ ---        ┆ ---        ┆ ---   ┆ ---    ┆ ---         │
#> │ date       ┆ f64        ┆ str   ┆ f64    ┆ str         │
#> ╞════════════╪════════════╪═══════╪════════╪═════════════╡
#> │ 2016-05-12 ┆ 82.19      ┆ b     ┆ 4164.0 ┆ a           │
#> │ 2017-05-12 ┆ 82.66      ┆ b     ┆ 4411.0 ┆ a           │
#> │ 2018-05-12 ┆ 83.12      ┆ a     ┆ 4566.0 ┆ b           │
#> │ 2019-05-12 ┆ 83.52      ┆ a     ┆ 4696.0 ┆ b           │
#> └────────────┴────────────┴───────┴────────┴─────────────┘

# .... and forward
pop$join_asof(gdp, on = "date", strategy = "forward")$collect()

#> shape: (4, 5)
#> ┌────────────┬────────────┬───────┬────────┬─────────────┐
#> │ date       ┆ population ┆ group ┆ gdp    ┆ group_right │
#> │ ---        ┆ ---        ┆ ---   ┆ ---    ┆ ---         │
#> │ date       ┆ f64        ┆ str   ┆ f64    ┆ str         │
#> ╞════════════╪════════════╪═══════╪════════╪═════════════╡
#> │ 2016-05-12 ┆ 82.19      ┆ b     ┆ 4411.0 ┆ a           │
#> │ 2017-05-12 ┆ 82.66      ┆ b     ┆ 4566.0 ┆ b           │
#> │ 2018-05-12 ┆ 83.12      ┆ a     ┆ 4696.0 ┆ b           │
#> │ 2019-05-12 ┆ 83.52      ┆ a     ┆ null   ┆ null        │
#> └────────────┴────────────┴───────┴────────┴─────────────┘

# join by a group: "only look within groups"
pop$join_asof(gdp, on = "date", by = "group", strategy = "backward")$collect()

#> shape: (4, 4)
#> ┌────────────┬────────────┬───────┬────────┐
#> │ date       ┆ population ┆ group ┆ gdp    │
#> │ ---        ┆ ---        ┆ ---   ┆ ---    │
#> │ date       ┆ f64        ┆ str   ┆ f64    │
#> ╞════════════╪════════════╪═══════╪════════╡
#> │ 2016-05-12 ┆ 82.19      ┆ b     ┆ 4321.0 │
#> │ 2017-05-12 ┆ 82.66      ┆ b     ┆ 4321.0 │
#> │ 2018-05-12 ┆ 83.12      ┆ a     ┆ 4411.0 │
#> │ 2019-05-12 ┆ 83.52      ┆ a     ┆ 4411.0 │
#> └────────────┴────────────┴───────┴────────┘

# only look 2 weeks and 2 days back
pop$join_asof(gdp, on = "date", strategy = "backward", tolerance = "2w2d")$collect()

#> shape: (4, 5)
#> ┌────────────┬────────────┬───────┬────────┬─────────────┐
#> │ date       ┆ population ┆ group ┆ gdp    ┆ group_right │
#> │ ---        ┆ ---        ┆ ---   ┆ ---    ┆ ---         │
#> │ date       ┆ f64        ┆ str   ┆ f64    ┆ str         │
#> ╞════════════╪════════════╪═══════╪════════╪═════════════╡
#> │ 2016-05-12 ┆ 82.19      ┆ b     ┆ null   ┆ null        │
#> │ 2017-05-12 ┆ 82.66      ┆ b     ┆ 4411.0 ┆ a           │
#> │ 2018-05-12 ┆ 83.12      ┆ a     ┆ null   ┆ null        │
#> │ 2019-05-12 ┆ 83.52      ┆ a     ┆ null   ┆ null        │
#> └────────────┴────────────┴───────┴────────┴─────────────┘

# only look 11 days back (numeric tolerance depends on polars type, <date> is in days)
pop$join_asof(gdp, on = "date", strategy = "backward", tolerance = 11)$collect()

#> shape: (4, 5)
#> ┌────────────┬────────────┬───────┬────────┬─────────────┐
#> │ date       ┆ population ┆ group ┆ gdp    ┆ group_right │
#> │ ---        ┆ ---        ┆ ---   ┆ ---    ┆ ---         │
#> │ date       ┆ f64        ┆ str   ┆ f64    ┆ str         │
#> ╞════════════╪════════════╪═══════╪════════╪═════════════╡
#> │ 2016-05-12 ┆ 82.19      ┆ b     ┆ null   ┆ null        │
#> │ 2017-05-12 ┆ 82.66      ┆ b     ┆ 4411.0 ┆ a           │
#> │ 2018-05-12 ┆ 83.12      ┆ a     ┆ null   ┆ null        │
#> │ 2019-05-12 ┆ 83.52      ┆ a     ┆ null   ┆ null        │
#> └────────────┴────────────┴───────┴────────┴─────────────┘