srlearn.Database

class srlearn.Database

Database of examples and facts.

__init__()

Initialize a Database object

A database (in this respect) contains positive examples, negative examples, facts, and is augmented with background knowledge.

The implementation is done with four attributes: pos, neg, facts, and modes. Each attribute is a list that may be set by mutating, or loaded from files with Database.from_files().

Examples

This initializes a Database object, then sets the pos attribute.

>>> from srlearn import Database
>>> db = Database()
>>> db.pos = ["student(alexander)."]

static from_files(pos='pos.pl', neg='neg.pl', facts='facts.pl', lazy_load=True)

Load files into a Database

Return an instance of a Database with pos, neg, and facts set to the contents of files. By default this performs a “lazy load,” where the files are not loaded into Python lists, but copied at learning time.

Parameters:

pos : str or pathlib.Path

Location of positive examples

neg : str or pathlib.Path

Location of negative examples

facts : str or pathlib.Path

Location of facts

lazy_load : bool (default: True)

Skip loading the files into a list

Returns:

db : srlearn.Database

Instance of a Database object

write(filename='train', location=PosixPath('train')) → None

Write the database to disk

Parameters:

filename : str

Name of the file to write to: ‘train’ or ‘test’

location : pathlib.Path

Path where data should be written to.

Notes

This function has polymorphic behavior. When attributes (self.pos, self.neg, self.facts) are lists of strings, the lists are written to files. When the attributes are (path-like) strings or pathlib Paths (pathlib.Path), the files are copied.

Examples using srlearn.Database

Smokes-Friends-Cancer

Family Relationships Domain