djutils
provides helpful tools for creating datajoint
pipelines.
pip3 install git+https://github.com/cajal/djutils.git
Standard lookup
and computed
datajoint tables can be created as follows:
from djutils import Schema
schema = Schema(...)
@schema.lookup
class ExampleLookup:
definition = """
...
"""
@schema.computed
class ExampleComputed:
definition = """
...
"""
def make(self, key):
...
In addition, djutils
provides table designs that are useful for creating composable and row-oriented pipelines that depend on abstractions, not concretions.
@schema.link
class Abstract:
links = [ConcreteTableA, ConcreteTableB, ...]
name = "abstract"
@schema.set
class ExampleSet:
keys = [TableA, TableB, ...]
name = "exampleset"
@schema.list
class ExampleList:
keys = [TableA, TableB, ...]
name = "examplelist"
... and much more.