Getting started

Using csv files

Here is a simple example which loads the famous East-West challenge dataset (trains) from two .csv files using CSVConnection. CSV files must contain two additional header lines beside column names which specify data types using SQLite syntax, and constraints such as primary and foreign key using simplified SQLite syntax. The

trains.csv

id	direction
integer	varchar
primary key
1	east
2	east
…	…
19	west
20	west

cars.csv

id	tid	position	shape	len	sides	roof	wheels	load_shape	load_num
integer	integer	integer	varchar	varchar	varchar	varchar	integer	varchar	integer
primary key	foreign key [trains.id]
1	1	1	rectangle	short	not_double	none	2	circle	1
2	1	2	rectangle	long	not_double	none	3	hexagon	1
3	1	3	rectangle	short	not_double	peaked	2	triangle	1
…	…	…	…	…	…	…	…	…	…
62	20	1	rectangle	long	not_double	flat	3	hexagon	1
63	20	2	u_shaped	short	not_double	none	2	triangle	1

from rdm.db import DBVendor, DBContext, AlephConverter, SQLiteDBConnection, CSVConnection
from rdm.wrappers import Aleph

# Provide data
connection = CSVConnection(['examples/data/trains/trains.csv',
                             'examples/data/trains/cars.csv'])

# Define learning context
context = DBContext(connection, target_table='trains', target_att='direction')

# Convert the data and induce features using Aleph
conv = AlephConverter(context, target_att_val='east')
aleph = Aleph()
theory, features = aleph.induce('induce_features', conv.positive_examples(),
                                conv.negative_examples(),
                                conv.background_knowledge())
print(theory)

Using SQLite

If the third line of the code above is changed as shown below the SQLite database containing the same data will be used.

connection = SQLiteDBConnection('examples/data/trains/trains.sqlite3')

Using MySQL

The same example but the data is now loaded from a MySQL database using DBConnection.

from rdm.db import DBVendor, DBConnection, DBContext, AlephConverter
from rdm.wrappers import Aleph

# Provide connection information
connection = DBConnection(
    'ilp',             # User
    'ilp123',          # Password
    'workflow.ijs.si', # Host
    'ilp',             # Database
)

# Define learning context
context = DBContext(connection, target_table='trains', target_att='direction')

# Convert the data and induce features using Aleph
conv = AlephConverter(context, target_att_val='east')
aleph = Aleph()
theory, features = aleph.induce('induce_features', conv.positive_examples(),
                                conv.negative_examples(),
                                conv.background_knowledge())
print(theory)

CSV files vs. RDBMS

Please note that not all wrapped ILP systems are able to use CSV files and SQLite databases. For example, Proper expects credentials to access the database directly.