https://github.com/tpapp/flexdates.jl
Dates with arbitrary integer types and epochs.
https://github.com/tpapp/flexdates.jl
Last synced: 3 months ago
JSON representation
Dates with arbitrary integer types and epochs.
- Host: GitHub
- URL: https://github.com/tpapp/flexdates.jl
- Owner: tpapp
- License: other
- Created: 2017-10-18T13:45:34.000Z (over 8 years ago)
- Default Branch: master
- Last Pushed: 2020-12-20T09:14:55.000Z (over 5 years ago)
- Last Synced: 2026-01-20T20:04:51.372Z (5 months ago)
- Language: Julia
- Size: 21.5 KB
- Stars: 3
- Watchers: 1
- Forks: 1
- Open Issues: 2
-
Metadata Files:
- Readme: README.md
- License: LICENSE.md
Awesome Lists containing this project
README
# FlexDates
[](https://travis-ci.org/tpapp/FlexDates.jl)
[](https://coveralls.io/github/tpapp/FlexDates.jl?branch=master)
[](http://codecov.io/github/tpapp/FlexDates.jl?branch=master)
## Motivation
Julia `Date`s in the standard library `Dates` represent a particular day counting from `0001-01-01` as the first day, using `Int64`. This allows the user to use dates in the interval of ±2.5e16 years, starting well before the Big Bang, and continuing way after the sun dies.
For the modest needs of social science data, a more limited range is
often sufficient. For example, `Int16` can represent days spanning a
bit more than 179 years (`2^16/365`), while `Int32` works for a bit
more than 10 million years.
This package helps with the management of dates, relying on `Dates` to do most of the tricky bits of date calculations. The main type in this package, `FlexDate`, is intended primarily as a *storage* format: using more compact dates, you can economize on memory and disk consumption.
## Installation
The package is not (yet) registered. Install with
```julia
pkg> add https://github.com/tpapp/FlexDates.jl
```
## Usage
You need to choose an *epoch*, a particular date we count from. This is incorporated into the type: Julia's zero cost abstraction ensures that there is no storage overhead when values are unboxed (eg arrays). The second type parameter for `FlexDate` is the integer type for internal representation. Constructors accept dates, `year, month, day`, or a single integer counting the number of days.
The following constructors are equivalent:
```julia
using FlexDates
julia> FlexDate{Date(2000,1,1), Int16}(Date(1980, 1, 1))
1980-01-01 [2000-01-01 + Int16 days]
julia> FlexDate{Date(2000,1,1), Int16}(1980, 1, 1)
1980-01-01 [2000-01-01 + Int16 days]
julia> FlexDate{Date(2000,1,1)}(Int16(-7305))
1980-01-01 [2000-01-01 + Int16 days]
```
When working with a dataset, it is recommended that you define a type constant:
```julia
const MyDate = FlexDate{Date(2000,1,1), Int16}
MyDate(1980, 1, 1)
```
Limited arithmetic and comparisons are supported. Keep in mind that these operations will be fast when you use dates with the *same epoch*, otherwise dates will be converted back to `Date` for calculations. A single dataset should use a single epoch unless there is a compelling reason to do otherwise.
For sophisticated arithmetic and date calculations, convert to `Date`, since `FlexDate` is meant primarily as a storage format.
For timespans, it is recommended that you use
[DiscreteRanges.jl](https://github.com/tpapp/DiscreteRanges.jl):
```julia
julia> using DiscreteRanges
julia> MyDate(2003, 4, 8) ∈ MyDate(2001, 1, 1)..MyDate(2010, 1, 1)
true
```