def selection(xs: List[str]) -> str:
xs_ = strip(xs)
return xs_[np.sum(editMatrix(xs_), axis=1).argmax()]
def editMatrix(xs: List[str]) -> Array:
return np.array([[levenshtein(x, y) for y in xs] for x in strip(xs)])
def selection(xs: List[str]) -> str:
xs_ = strip(xs)
return label(xs_)
def condition(xs: List[str]) -> float:
"average edit distance on cluster"
xs_ = strip(xs)
return (0.0 # levenshtein returns 1 if they are identical
if len(xs_) <= 1 else np.mean(
[levenshtein(*ys) for ys in itertools.combinations(xs_, 2)]))
def condition(xs: List[str]) -> float:
"longest common ngram"
xs_ = strip(xs)
return (0.0 if ((len(xs_) <= 1) or (np.min([len(x) for x in xs_]) < 3))
#TODO: if there is only one item then its vacuously 0?
else scorer(xs_))
# and that's it so far for generics, but we have a lot more list operations
# and they revolve around the structure of your data types
# mapmap: for when you have multiple layers of structure to map a map over
mapmap(lambda x: x**2, [[1,2], [2,3,4]])
# unpack: make a list of lists into a list
# unpackG: as unpack but return a generator expression
unpack([[1,2], [2,3,4]])
for i in unpackG([[1,2], [2,3,4]]):
print(i)
# strip: remove values from a list which evaluate to False
strip(['hi', '', 0, 'pepeHands'])
# all of these functions really form the backbone of how woffle was built and
# find use across a large swathe of the code
# if you think woffle may be of use to you or if you would like to add to it
# then you need to think about the steps in the workflow, if you can break it
# down into the 4 tasks mentioned then please PR and get things added!
# There is already a train directory in the hierarchy but there is nothing in
# it yet, if this is something you are interested in please PR and get it
# added.
# That is all for today, the examples you have seen are available in the
# examples directory of the repo along with a much more in depth re-working of
# optimus, pull it and give it a go!
def condition(xs: List[str]) -> float:
"implement hypernym lookup"
xs_ = strip(xs)
return (0.0 if len(xs_) <= 1 else 1.0)
def condition(xs: List[str]) -> float:
"average number of common words across the cluster"
xs_ = strip(xs)
return (0.0 if (len(xs_) <= 1) or (wordLen(xs_) < 3) else scorer(xs_))
def test_strip(xs):
assert lst.strip(xs) == [x for x in xs if x]