kurtbrose

m2m = M2M()
ch = M2MChain()
gr = M2MGraph()


# TRANSFORMING BETWEEN TYPES

chain(*ch.m2ms)  # kind of a copy
chain(ch)  # also results in copy

kurtbrose

import collections
import functools


def log_methods(cls):
    def __init__(self, *a, **kw):
        cls.__init__(self, *a, **kw)
        self._log = collections.deque(maxlen=100)

    def logit(method):

kurtbrose

"""
experimenting with increasingly general notions of indexing
"""
from relativity import M2M


class IndexedTriple(object):
    """
    keeps track of "left" to "right" pairs
    """

kurtbrose

>>> import timeit
>>> LIST = [None] * 10
>>> countrange = lambda: [1 for i in range(len(LIST))]
>>> countdirect = lambda: [1 for e in LIST]
>>> timeit.timeit(countrange, number=1000) * 1000
0.8033999999952357
>>> timeit.timeit(countdirect, number=1000) * 1000
0.5680000000012342
>>> LIST = [None] * 100
>>> timeit.timeit(countrange, number=1000) * 1000

kurtbrose

"""
A Schema is a more general container for Python objects.

In addition to tracking relationships between objects,
it can also keep other kind of indexing structures.

Will need a more flexible query object that can represent
comparisons.

NOTE -- not sure if this is really worth it since it only

kurtbrose

import glom


class F(object):
    """
    F for filter!

    An F object encapsulates boolean comparisons in a similar
    manner to SQLAlchemy or Pandas columns.

kurtbrose

_PAIRING = object()  # marker


#TODO: rename to Relation
class ManyToMany(object):
    """
    a dict-like entity that represents a many-to-many relationship
    between two groups of objects

    behaves like a dict-of-tuples; also has .inv which is kept

kurtbrose

import attr
import numpy as np

class TimeSampler(object)
    '''
    samples from a time indexed series based on time -- pick a random
    time within the range of the series; returns the time-delta
    and value-delta of the point to the right of the sample with the
    point before it
    

kurtbrose

kurtbrose

'''
line interpolation class for doing a linear regression
'''
import numpy as np
import pandas
import attr


@attr.s
class Line(object):