New release: 2.0.5!
pip install cogdb
Cog is a persistent embedded graph database implemented purely in Python. Torque is Cog's graph query language. Cog also provides a low level API to its fast persistent key-value store.
Cog is ideal for python applications that does not require a full featured database. Cog can easily be used as a library from within a Python application. Cog be used interactively in an IPython environment like Jupyter notebooks.
Cog can load a graph stored as N-Triples, a serialization format for RDF. See Wikipedia, W3C for details.
In short, an N-Triple is sequence of subject, predicate and object in a single line that defines a connection between two vertices:
vertex <predicate> vertex
from cog.torque import Graph
g = Graph("people")
g.put("alice","follows","bob")
g.put("bob","follows","fred")
g.put("bob","status","cool_person")
g.put("charlie","follows","bob")
g.put("charlie","follows","dani")
g.put("dani","follows","bob")
g.put("dani","follows","greg")
g.put("dani","status","cool_person")
g.put("emily","follows","fred")
g.put("fred","follows","greg")
g.put("greg","status","cool_person")
from cog.torque import Graph
g = Graph("books")
g.load_csv('test/test-data/books.csv', "isbn")
g.scan(3)
{'result': [{'id': 'bob'}, {'id': 'emily'}, {'id': 'charlie'}]}
g.scan(3, 'e')
{'result': [{'id': 'status'}, {'id': 'follows'}]}
g.v("bob").out().all()
{'result': [{'id': 'cool_person'}, {'id': 'fred'}]}
g.v().has("status", 'cool_person').all()
{'result': [{'id': 'bob'}, {'id': 'dani'}, {'id': 'greg'}]}
g.v().has("follows", "fred").inc().all('e')
{'result': [{'id': 'dani', 'edges': ['follows']}, {'id': 'charlie', 'edges': ['follows']}, {'id': 'alice', 'edges': ['follows']}]}
g.v("bob").out().count()
'2'
Note: render()
is supported only in IPython environment like Jupyter notebook otherwise use view(..).url.
By tagging the vertices 'from' and 'to', the resulting graph can be visualized.
g.v().tag("from").out("follows").tag("to").view("follows").render()
g.v().tag("from").out("follows").tag("to").view("follows").url
file:///Path/to/your/cog_home/views/follows.html
g.lsv()
['follows']
g.getv('follows').render()
g.v("bob").out().tag("from").out().tag("to").all()
{'result': [{'from': 'fred', 'id': 'greg', 'to': 'greg'}]}
g.v("bob").inc().all()
{'result': [{'id': 'alice'}, {'id': 'charlie'}, {'id': 'dani'}]}
from cog.torque import Graph
g = Graph(graph_name="people")
g.load_triples("/path/to/triples.nt", "people")
from cog.torque import Graph
g = Graph(graph_name="people")
g.load_edgelist("/path/to/edgelist", "people")
Every record inserted into Cog's key-value store is directly persisted on to disk. It stores and retrieves data based on hash values of the keys, it can perform fast look ups (O(1) avg) and fast (O(1) avg) inserts.
from cog.database import Cog
cogdb = Cog('path/to/dbdir')
# create a namespace
cogdb.create_namespace("my_namespace")
# create new table
cogdb.create_table("new_db", "my_namespace")
# put some data
cogdb.put(('key','val'))
# retrieve data
cogdb.get('key')
# put some more data
cogdb.put(('key2','val2'))
# scan
scanner = cogdb.scanner()
for r in scanner:
print r
# delete data
cogdb.delete('key1')
If no config is provided when creating a Cog instance, it will use the defaults:
COG_PATH_PREFIX = "/tmp"
COG_HOME = "cog-test"
from cog import config
config.COG_HOME = "app1_home"
data = ('user_data:id=1','{"firstname":"Hari","lastname":"seldon"}')
cog = Cog(config)
cog.create_namespace("test")
cog.create_table("db_test", "test")
cog.put(data)
scanner = cog.scanner()
for r in scanner:
print r
Put and Get calls performance:
put ops/second: 18968
get ops/second: 39113
The perf test script is included with the tests: insert_bench.py and get_bench.py
INDEX_LOAD_FACTOR on an index determines when a new index file is created, Cog uses linear probing to resolve index collisions. Higher INDEX_LOAD_FACTOR leads slightly lower performance on operations on index files that have reached the target load.