Duration

A time argument computed as the elapsed time between two date or datetime columns.

Usage

Duration(
    start,
    end,
    unit="days",
)

A Duration is a frozen specification that names two date or datetime columns and a time unit. When passed to an Outcome constructor as the time (or time2) argument, Greenwood computes the elapsed time automatically when the data is read. Inside a formula, write the same thing as duration(start, end, unit="days").

Create a Duration with the duration() helper rather than constructing it directly.

Parameters

start: str: Column name holding the origin date or datetime (e.g., enrollment date).
end: str: Column name holding the exit date or datetime (e.g., date of event or last contact).
unit: str = "days": Time unit for the elapsed duration: "days" (default), "weeks", "months", "years", "hours", "minutes", or "seconds". Months and years use average Gregorian lengths (30.4375 and 365.25 days).

Examples

Create a Duration that computes follow-up time in days between two date columns, and use it as the time argument of an Outcome:

import greenwood as gw
import polars as pl

df = pl.DataFrame({
    "enroll": ["2020-01-01", "2020-03-15", "2020-06-01"],
    "last_contact": ["2021-06-15", "2020-12-01", "2022-01-10"],
    "event": [1, 0, 1],
}).with_columns(
    pl.col("enroll").str.to_date(),
    pl.col("last_contact").str.to_date(),
)

d = gw.duration(start="enroll", end="last_contact", unit="days")
d

duration(start='enroll', end='last_contact', unit='days')

y = gw.Outcome.surv(time=d, event="event").bind(df)
y

Surv(type=right, n=3, events=2)

Attributes

Name	Description
columns	The two columns the duration reads.

columns

The two columns the duration reads.

columns: tuple[str, str]

Methods

Name	Description
compute()	The elapsed time from start to `end` in `unit` (`NaN` where either is missing).

compute()

The elapsed time from start to end in unit (NaN where either is missing).

Usage

Source

compute(
    start,
    end,
)