clothesline

A library to handle sets, made of intervals over a continuous infinite axis (a "domain") such as the real numbers or date/times.

import clothesline
bld = clothesline.RealIntervalSet.builder()
set1 = bld[-10](10) - bld(0)[1]
set1                       # [-10, 0] U (1, 10)
set1.extension()           # 19
set2 = set1 + bld[100](...)
set2.complement()          # (-inf, -10) U (0, 1] U [10, 100)
set2 + set2.complement()   # (-inf, +inf)

Quickstart

Note: for the full documentation, visit clothesline.readthedocs.io.

Work with sets

The library provides ready-to-use RealIntervalSet to deal with sets over a number line and DatetimeIntervalSet for date intervals.

Start by importing the library and getting a builder for interval sets on the real line:

import clothesline
bld = clothesline.RealIntervalSet.builder()

Create a couple of intervals:

set1 = bld(0)(10)
set2 = bld[0][1] + bld(2)(3) + bld[20](...)
print(set2)

Here, round brackets mean that the boundary is excluded; square brackets mean the point is included. Three dots stand for the point at infinity.

You can also get an utils object, to create standard sets such as open or closed intervals, point-like sets and so on:

uti = clothesline.RealIntervalSet.utils()
set3 = uti.open(0, 5)
set4 = uti.high_slice(10, included=True)

Most set operations are supported: from complement to difference, from inclusion tests to XOR. Interval Sets should be treated as immutable objects. Moreover, since the internal representation of a set is always normalized to a canonical form, equality tests work as you would expect:

set5 = set4.union(set2)
set4 - set2
set3 + bld[3](10) == set1   # True

For more ways to create and manipulate sets, have a look at the User Guide in the full documentation.

Store and retrieve

You can convert any interval set into a string (e.g. to store it into a database) and re-hydrate it later:

import json
serialized = json.dumps(set3.to_dict())
# ...
new_set3 = uti.from_dict(json.loads(serialized))
new_set3 == set3    # True

Datetimes

Support for datetimes is ready out-of-the-box:

from datetime import datetime
dbld = clothesline.DatetimeIntervalSet.builder()
timespan1 = dbld[datetime(2010, 10, 10)](datetime(2011, 2, 20))
print(timespan1.complement())
print(timespan1.extension())

That's it! For more, check the full documentation.

Contributing

Coming soon.