def pandas_read_csv(self, usecols=None, **kwargs):
""" Use pandas.read_csv with the right keyword arguments
In particular we know what dtypes should be, which columns are dates,
etc...
"""
dtypes, dates = dshape_to_pandas(self.schema)
if usecols:
if builtins.all(isinstance(c, int) for c in usecols):
usecols = get(usecols, self.columns)
dates = [name for name in dates if name in usecols]
result = pd.read_csv(self.path,
names=kwargs.pop('names', self.columns),
usecols=usecols,
compression={'gz': 'gzip',
'bz2': 'bz2'}.get(ext(self.path)),
dtype=kwargs.pop('dtype', dtypes),
parse_dates=kwargs.pop('parse_dates', dates),
encoding=kwargs.pop('encoding', self.encoding),
header=0 if self.header else None,
**merge(kwargs, clean_dialect(self.dialect)))
reorder = get(list(usecols)) if usecols and len(usecols) > 1 else identity
if isinstance(result, (pd.Series, pd.DataFrame)):
return reorder(result)
else:
return map(reorder, result)
def test_markov_tables():
markov_features = ['markov_N', 'markov_R', 'markov_NR', 'markov_RN', 'markov_NN', 'markov_RR']
assert(t.get(markov_features, f.markov_tables(markovtesttree), str(0.0)) ==
('0.809523809524', '0.190476190476', '0.285714285714', '0.285714285714', '0.428571428571', '0.0'))
assert(t.get(markov_features, f.markov_tables(smalltree), str(0.0)) ==
('1.0', '0.0', '0.0', '0.0', '1.0', '0.0'))
assert(t.get(markov_features, f.markov_tables(f.compress(smalltree)), str(0.0)) ==
('1.0', '0.0', '0.0', '0.0', '0.0', '0.0'))
def select_permits():
types = get(permit_type_checkbox.active, permit_types)
res_non = get(res_non_checkbox.active, res_non_types)
selected = df[(df.year >= min_year.value) &
(df.year <= max_year.value) &
(df.permit_value >= min_permit_cost.value) &
(df.permit_value <= max_permit_cost.value) &
(df.type.isin(types)) &
(df.res_non.isin(res_non))]
return selected
def records_to_tuples(ds, data):
""" Transform records into tuples
Examples
--------
>>> seq = [{'a': 1, 'b': 10}, {'a': 2, 'b': 20}]
>>> list(records_to_tuples('var * {a: int, b: int}', seq))
[(1, 10), (2, 20)]
>>> records_to_tuples('{a: int, b: int}', seq[0]) # single elements
(1, 10)
>>> records_to_tuples('var * int', [1, 2, 3]) # pass through on non-records
[1, 2, 3]
See Also
--------
tuples_to_records
"""
if isinstance(ds, (str, unicode)):
ds = dshape(ds)
if isinstance(ds.measure, Record) and len(ds.shape) == 1:
return pluck(ds.measure.names, data, default=None)
if isinstance(ds.measure, Record) and len(ds.shape) == 0:
return get(ds.measure.names, data)
if not isinstance(ds.measure, Record):
return data
raise NotImplementedError()
def partial_reduce(func, x, split_every, keepdims=False, dtype=None, name=None):
"""Partial reduction across multiple axes.
Parameters
----------
func : function
x : Array
split_every : dict
Maximum reduction block sizes in each dimension.
Example
-------
Reduce across axis 0 and 2, merging a maximum of 1 block in the 0th
dimension, and 3 blocks in the 2nd dimension:
>>> partial_reduce(np.min, x, {0: 1, 2: 3}) # doctest: +SKIP
"""
name = name or 'p_reduce-' + tokenize(func, x, split_every, keepdims, dtype)
parts = [list(partition_all(split_every.get(i, 1), range(n))) for (i, n)
in enumerate(x.numblocks)]
keys = product(*map(range, map(len, parts)))
out_chunks = [tuple(1 for p in partition_all(split_every[i], c)) if i
in split_every else c for (i, c) in enumerate(x.chunks)]
if not keepdims:
out_axis = [i for i in range(x.ndim) if i not in split_every]
getter = lambda k: get(out_axis, k)
keys = map(getter, keys)
out_chunks = list(getter(out_chunks))
dsk = {}
for k, p in zip(keys, product(*parts)):
decided = dict((i, j[0]) for (i, j) in enumerate(p) if len(j) == 1)
dummy = dict(i for i in enumerate(p) if i[0] not in decided)
g = lol_tuples((x.name,), range(x.ndim), decided, dummy)
dsk[(name,) + k] = (func, g)
return Array(merge(dsk, x.dask), name, out_chunks, dtype=dtype)
def post_compute(e, q, d):
"""
Execute a query using MongoDB's aggregation pipeline
The compute_up functions operate on Mongo Collection / list-of-dict
queries. Once they're done we need to actually execute the query on
MongoDB. We do this using the aggregation pipeline framework.
http://docs.mongodb.org/manual/core/aggregation-pipeline/
"""
d = {'$project': toolz.merge({'_id': 0}, # remove mongo identifier
dict((col, 1) for col in e.fields))}
q = q.append(d)
if not e.dshape.shape: # not a collection
result = q.coll.aggregate(list(q.query))['result'][0]
if isscalar(e.dshape.measure):
return result[e._name]
else:
return get(e.fields, result)
dicts = q.coll.aggregate(list(q.query))['result']
if isscalar(e.dshape.measure):
return list(pluck(e.fields[0], dicts, default=None)) # dicts -> values
else:
return list(pluck(e.fields, dicts, default=None)) # dicts -> tuples
def finalize(bases):
shape = bases[0].shape[:2]
out = nd.empty(shape, dshape)
for path, finalizer, inds in zip(paths, finalizers, indices):
arr = reduce(getattr, path, out)
np_arr = nd.as_numpy(arr.view_scalars(arr.dtype.value_type))
np_arr[:] = finalizer(*get(inds, bases))
return out
def __init__(self, lhs, rhs, on_left=None, on_right=None):
self.lhs = lhs
self.rhs = rhs
if not on_left and not on_right:
on_left = on_right = unpack(list(sorted(
set(lhs.columns) & set(rhs.columns),
key=lhs.columns.index)))
if not on_right:
on_right = on_left
if isinstance(on_left, tuple):
on_left = list(on_left)
if isinstance(on_right, tuple):
on_right = list(on_right)
self._on_left = tuple(on_left) if isinstance(on_left, list) else on_left
self._on_right = (tuple(on_right) if isinstance(on_right, list)
else on_right)
if get(on_left, lhs.schema[0]) != get(on_right, rhs.schema[0]):
raise TypeError("Schema's of joining columns do not match")
def func(scheduler):
""" Get CPU and Memory usage on each worker """
workers = [k for k, v in sorted(scheduler.ncores.items(),
key=lambda x: x[0], reverse=True)]
nannies = [(ip, scheduler.nannies[(ip, port)])
for ip, port in workers]
dicts = [get(-1, scheduler.resource_logs[w], dict())
for w in nannies]
return {'workers': workers,
'cpu': [d.get('cpu_percent', -1) for d in dicts],
'memory': [d.get('memory_percent', -1) for d in dicts]}
def rget(sequence, key, default=None):
"""Get element in a sequence or dict.
Like toolz.get but with parameters in reverse order.
Args:
sequence (sequence or dict): sequence or dict
key (str or int): key to access in sequence
Returns:
object: value behind the key
"""
return get(key, sequence, default=default)
def join(lhs, rhs, on_left=None, on_right=None, how='inner'):
if not on_left and not on_right:
on_left = on_right = unpack(list(sorted(
set(lhs.columns) & set(rhs.columns),
key=lhs.columns.index)))
if not on_right:
on_right = on_left
if isinstance(on_left, tuple):
on_left = list(on_left)
if isinstance(on_right, tuple):
on_right = list(on_right)
if get(on_left, lhs.schema[0]) != get(on_right, rhs.schema[0]):
raise TypeError("Schema's of joining columns do not match")
_on_left = tuple(on_left) if isinstance(on_left, list) else on_left
_on_right = (tuple(on_right) if isinstance(on_right, list)
else on_right)
how = how.lower()
if how not in ('inner', 'outer', 'left', 'right'):
raise ValueError("How parameter should be one of "
"\n\tinner, outer, left, right."
"\nGot: %s" % how)
return Join(lhs, rhs, _on_left, _on_right, how)
def get_prox_key(infos, key, default=None, reduce=None, array=True):
""" Build array of prox output for each operator and iteration.
Return an {#iterations} by {#prox operators} array (unless a reduction is performed).
reduce is usually, np.mean, np.min, or np.max
Some prox outputs may be `None` or `{}`. In that case,
return the `default` value.
"""
g = ([get(key, p, default=default) for p in info['prox_infos']] for info in infos )
if reduce:
g = (reduce(row) for row in g)
g = list(g)
if array:
g = np.array(g)
return g
def _make_segregator(sles, groupby_filter, partitions):
groupby_fn = compose(
partial(get, "key", default=None),
excepts(StopIteration, first, lambda __: {}),
partial(flip, filter, partitions),
groupby_filter,
)
sles_grouped = groupby(groupby_fn, sles)
def seg_filter(x):
return lambda sl: sl.get("item_code") == x
summer = compose(operator.neg, sum, partial(pluck, "actual_qty"))
def seg_reducer(item_code):
def fn(a, p):
key = get("key", p, None)
seger = get("seger", p, lambda __: None)
return merge(a, {key: seger(item_code)})
return fn
segregator_fns = [
merge(
x,
{
"seger":
compose(
summer,
partial(flip, filter, get(x.get("key"), sles_grouped, [])),
seg_filter,
)
},
) for x in partitions
]
def fn(item_code):
return reduce(seg_reducer(item_code), segregator_fns, {})
return fn
def main(input_file, output_file):
reader = csv.reader(input_file)
writer = csv.DictWriter(
output_file,
fieldnames=["date", "latitude", "longitude"] +
listmap(second, WEATHER_FIELDS),
)
writer.writeheader()
for row in reader:
date, latitude, longitude, _, payload = row
weather = get_in(["daily", "data", 0], json.loads(payload), {})
writer.writerow({
"date": date,
"latitude": latitude,
"longitude": longitude,
**{
field_name: get(field, weather, None)
for field, field_name in WEATHER_FIELDS
},
})
def transform(example, output_size=(HEIGHT, WIDTH), training=True):
image, objects, image_id = decode(example)
# if training:
# image = random_resize(image, output_size, ratio_range=(0.8, 1.2))
# image, objects = random_crop(image, objects, output_size)
# image, objects = random_hflip(image, objects, 0.5)
# else:
image = resize(image, output_size)
image = normalize(image, [123.68, 116.779, 103.939], [58.393, 57.12, 57.375])
image, objects = pad_to(image, objects, output_size)
gt_bboxes, gt_labels = get(['gt_bbox', 'gt_label'], objects)
gt_bboxes = coords_to_absolute(gt_bboxes, tf.shape(image)[:2])
bbox_targets, labels, centerness = atss_match(
gt_bboxes, gt_labels, anchors, num_level_bboxes, topk=9, centerness=True)
return image, {'bbox_target': bbox_targets, 'label': labels, 'centerness': centerness,
'image_id': image_id}
def Slideshow(text):
'''
Shows an image with arrows to go to next/previous images.
---image-slider
url1::caption1
url2::caption2
...
Will make an image slider with those images in that order.
'''
lines = list(filter(lambda x: x.strip() != "", text.split('\n')))
name = f'ss-{int(datetime.now().timestamp())}-{re.sub(r"[^a-z]", "", lines[0])}'
return ['div.slideshow', {'data-transition': 'fade'}, *[
[['input.slideshow--bullet',
{'type': 'radio', 'name': name, 'id': f'{name}-item-{i}', 'checked': i == 0}],
['div.slideshow--item', {'data-pos': f"{i+1}/{len(lines)}"},
['img', {'src': line.split('::')[0]}],
['div.slideshow--caption', t.get(1, line.split('::'), '')],
['label.slideshow--nav.slideshow--nav-previous', {'for': f'{name}-item-{(i-1)%len(lines)}'}, f'Go to slide {(i-1)%len(lines) + 1}'],
['label.slideshow--nav.slideshow--nav-next', {'for': f'{name}-item-{(i+1)%len(lines)}'}, f'Go to slide {(i+1)%len(lines) + 1}']
]
] for (i, line) in enumerate(lines)]]
def partial_reduce(func, x, split_every, keepdims=False, dtype=None, name=None):
""" Partial reduction across multiple axes.
Parameters
----------
func : function
x : Array
split_every : dict
Maximum reduction block sizes in each dimension.
Examples
--------
Reduce across axis 0 and 2, merging a maximum of 1 block in the 0th
dimension, and 3 blocks in the 2nd dimension:
>>> partial_reduce(np.min, x, {0: 1, 2: 3}) # doctest: +SKIP
"""
name = (name or funcname(func)) + '-' + tokenize(func, x, split_every,
keepdims, dtype)
parts = [list(partition_all(split_every.get(i, 1), range(n))) for (i, n)
in enumerate(x.numblocks)]
keys = product(*map(range, map(len, parts)))
out_chunks = [tuple(1 for p in partition_all(split_every[i], c)) if i
in split_every else c for (i, c) in enumerate(x.chunks)]
if not keepdims:
out_axis = [i for i in range(x.ndim) if i not in split_every]
getter = lambda k: get(out_axis, k)
keys = map(getter, keys)
out_chunks = list(getter(out_chunks))
dsk = {}
for k, p in zip(keys, product(*parts)):
decided = dict((i, j[0]) for (i, j) in enumerate(p) if len(j) == 1)
dummy = dict(i for i in enumerate(p) if i[0] not in decided)
g = lol_tuples((x.name,), range(x.ndim), decided, dummy)
dsk[(name,) + k] = (func, g)
return Array(sharedict.merge(x.dask, (name, dsk), dependencies={name: {x.name}}),
name, out_chunks, dtype=dtype)
def make_envs(cost=1.00, n=100, seed=None, variance_structure="constant_high"):
if seed is not None:
np.random.seed(seed)
sigmas = get(
variance_structure, {
"constant_high": [0, 20, 20, 20],
"increasing": [0, 2, 4, 20],
"decreasing": [0, 20, 10, 5],
"constant_low": [0, 1, 1, 1],
})
def reward(depth):
if depth > 0:
return Normal(0, sigmas[depth]).to_discrete(6)
return 0.
branching = [4, 1, 2]
envs = [
MouselabEnv.new_symmetric(branching, reward, cost=cost)
for _ in range(n)
]
return envs
def types_of_fields(fields, expr):
""" Get the types of fields in an expression
Examples
--------
>>> from blaze import symbol
>>> expr = symbol('e', 'var * {x: int64, y: float32}')
>>> types_of_fields('y', expr)
ctype("float32")
>>> types_of_fields(['y', 'x'], expr)
(ctype("float32"), ctype("int64"))
>>> types_of_fields('x', expr.x)
ctype("int64")
"""
if isinstance(expr.dshape.measure, Record):
return get(fields, expr.dshape.measure)
else:
if isinstance(fields, (tuple, list, set)):
assert len(fields) == 1
fields, = fields
assert fields == expr._name
return expr.dshape.measure
def types_of_fields(fields, expr):
""" Get the types of fields in an expression
Examples
--------
>>> from blaze import symbol
>>> expr = symbol('e', 'var * {x: int64, y: float32}')
>>> types_of_fields('y', expr)
ctype("float32")
>>> types_of_fields(['y', 'x'], expr)
(ctype("float32"), ctype("int64"))
>>> types_of_fields('x', expr.x)
ctype("int64")
"""
if isinstance(expr.dshape.measure, Record):
return get(fields, expr.dshape.measure)
else:
if isinstance(fields, (tuple, list, set)):
assert len(fields) == 1
fields, = fields
assert fields == expr._name
return expr.dshape.measure
def onload(self):
all_fees = frappe.db.sql(
"""
SELECT
si.rounded_total AS amount,
fee.status AS status,
fee.to_date AS end_date
FROM `tabGym Fee` AS fee, `tabSales Invoice` AS si
WHERE
fee.docstatus = 1 AND
fee.membership = '{membership}' AND
fee.reference_invoice = si.name
ORDER BY fee.to_date DESC
""".format(membership=self.name),
as_dict=True,
)
unpaid_fees = filter(lambda x: x.get('status') == 'Unpaid', all_fees)
self.set_onload('total_invoices', count(all_fees))
self.set_onload('unpaid_invoices', count(unpaid_fees))
outstanding = reduce(operator.add, pluck('amount', unpaid_fees), 0)
self.set_onload('outstanding', outstanding)
paid_fees = filter(lambda x: x.get('status') == 'Paid', all_fees)
end_date = get('end_date', first(paid_fees)) if paid_fees else None
self.set_onload('end_date', end_date)
def preprocess(example,
output_size=(HEIGHT, WIDTH),
max_objects=50,
training=True):
image, objects, image_id = decode(example)
# if training:
# image = random_resize(image, output_size, ratio_range=(0.8, 1.2))
# image, objects = random_crop(image, objects, output_size)
# image, objects = random_hflip(image, objects, 0.5)
# else:
image = resize(image, output_size)
image = normalize(image, [123.68, 116.779, 103.939],
[58.393, 57.12, 57.375])
image, objects = pad_to(image, objects, output_size)
gt_bboxes, gt_labels = get(['gt_bbox', 'gt_label'], objects)
gt_bboxes = coords_to_absolute(gt_bboxes, tf.shape(image)[:2])
objects = {**objects, 'gt_bbox': gt_bboxes}
bbox_targets, labels, ignore = max_iou_match(gt_bboxes,
gt_labels,
bbox_coder,
pos_iou_thr=0.5,
neg_iou_thr=0.4,
encode_bbox=False)
objects = pad_objects(objects, max_objects)
return image, {
'bbox_target': bbox_targets,
'label': labels,
'ignore': ignore,
**objects, 'image_id': image_id
}
def get_pos_data():
from erpnext.accounts.doctype.sales_invoice.pos import get_pos_data
data = get_pos_data()
allowed_items = get("bin_data", data, {}).keys()
prices = _get_item_prices(allowed_items)
def set_prices(item):
get_price = compose(partial(get, seq=prices, default={}),
partial(get, "item_code"))
return merge(item, get_price(item))
trans_items = compose(
partial(map, set_prices),
partial(filter, lambda x: x.get("name") in allowed_items),
partial(get, "items", default=[]),
)
add_branch = compose(
flip(merge, {"os_branch": get_user_branch()}),
lambda x: x.as_dict(),
partial(get, "doc", default={}),
)
return merge(data, {"items": trans_items(data), "doc": add_branch(data)})
def preprocess(example,
output_size=(HEIGHT, WIDTH),
max_objects=50,
training=True):
image, objects, image_id = decode(example)
# if training:
# image = random_resize(image, output_size, ratio_range=(0.8, 1.2))
# image, objects = random_crop(image, objects, output_size)
# image, objects = random_hflip(image, objects, 0.5)
# else:
image = resize(image, output_size)
image = normalize(image, [123.68, 116.779, 103.939],
[58.393, 57.12, 57.375])
image, objects = pad_to(image, objects, output_size)
gt_bboxes, gt_labels = get(['gt_bbox', 'gt_label'], objects)
gt_bboxes = coords_to_absolute(gt_bboxes, tf.shape(image)[:2])
objects = {**objects, 'gt_bbox': gt_bboxes}
bbox_targets, labels, centerness = fcos_match(gt_bboxes,
gt_labels,
points,
num_level_points,
strides=strides,
radius=0.5)
objects = pad_objects(objects, max_objects)
return image, {
'bbox_target': bbox_targets,
'label': labels,
'centerness': centerness,
**objects, 'image_id': image_id
}
def partial_reduce(func,
x,
split_every,
keepdims=False,
dtype=None,
name=None,
reduced_meta=None):
""" Partial reduction across multiple axes.
Parameters
----------
func : function
x : Array
split_every : dict
Maximum reduction block sizes in each dimension.
Examples
--------
Reduce across axis 0 and 2, merging a maximum of 1 block in the 0th
dimension, and 3 blocks in the 2nd dimension:
>>> partial_reduce(np.min, x, {0: 1, 2: 3}) # doctest: +SKIP
"""
name = (name or funcname(func)) + '-' + tokenize(func, x, split_every,
keepdims, dtype)
parts = [
list(partition_all(split_every.get(i, 1), range(n)))
for (i, n) in enumerate(x.numblocks)
]
keys = product(*map(range, map(len, parts)))
out_chunks = [
tuple(1 for p in partition_all(split_every[i], c))
if i in split_every else c for (i, c) in enumerate(x.chunks)
]
if not keepdims:
out_axis = [i for i in range(x.ndim) if i not in split_every]
getter = lambda k: get(out_axis, k)
keys = map(getter, keys)
out_chunks = list(getter(out_chunks))
dsk = {}
for k, p in zip(keys, product(*parts)):
decided = dict((i, j[0]) for (i, j) in enumerate(p) if len(j) == 1)
dummy = dict(i for i in enumerate(p) if i[0] not in decided)
g = lol_tuples((x.name, ), range(x.ndim), decided, dummy)
dsk[(name, ) + k] = (func, g)
graph = HighLevelGraph.from_collections(name, dsk, dependencies=[x])
meta = x._meta
if reduced_meta is not None:
try:
meta = func(reduced_meta, meta=True)
# no meta keyword argument exists for func, and it isn't required
except TypeError:
meta = func(reduced_meta)
# when no work can be computed on the empty array (e.g., func is a ufunc)
except ValueError:
pass
# some functions can't compute empty arrays (those for which reduced_meta
# fall into the ValueError exception) and we have to rely on reshaping
# the array according to len(out_chunks)
if not np.isscalar(meta) and meta.ndim != len(out_chunks):
if len(out_chunks) == 0:
meta = meta.sum()
else:
meta = meta.reshape((0, ) * len(out_chunks))
if np.isscalar(meta):
return Array(graph, name, out_chunks, dtype=dtype)
else:
return Array(graph, name, out_chunks, meta=meta.astype(dtype))
skip_prefetch=True))
it = iter(ds_train)
example = next(it)
image, objects, image_id = decode(example)
# if training:
# image = random_resize(image, output_size, ratio_range=(0.8, 1.2))
# image, objects = random_crop(image, objects, output_size)
# image, objects = random_hflip(image, objects, 0.5)
# else:
image = resize(image, output_size)
image = normalize(image, [123.68, 116.779, 103.939], [58.393, 57.12, 57.375])
image, objects = pad_to(image, objects, output_size)
gt_bboxes, gt_labels = get(['gt_bbox', 'gt_label'], objects)
gt_bboxes = coords_to_absolute(gt_bboxes, tf.shape(image)[:2])
objects = {**objects, 'gt_bbox': gt_bboxes}
bbox_targets, labels, centerness = atss_match(gt_bboxes,
gt_labels,
anchors,
num_level_bboxes,
topk=9,
centerness=True)
objects = pad_objects(objects, max_objects)
def combine(base_tuples):
bases = tuple(np.stack(bs) for bs in zip(*base_tuples))
return tuple(f(*get(inds, bases)) for (f, inds) in calls)
def _devices(attrs):
return [
'{}:{}'.format(*get(['PathOnHost', 'PathInContainer'], dev))
for dev in attrs
]
def __getitem__(self, key):
if isinstance(key, list):
return RecordAggregate(dict(zip(key, get(key, self._data))),
self.x_axis, self.y_axis)
return self._data[key]
def fn(row):
mops = get(row.get("sales_invoice"), payments_grouped, {})
return merge(row, mops, {"total_collected": sum(mops.values())})
def seger(sum_fn, x):
return compose(
sum_fn,
partial(flip, filter, get(x.get("key"), sales_grouped, [])),
seg_filter,
)
def finalize(bases, **kwargs):
data = {key: finalizer(get(inds, bases), **kwargs)
for (key, finalizer, inds) in calls}
return RecordAggregate(data, **kwargs)
def test_get():
for tup in tuples:
get([1, 2], tup)
def finalize(bases, **kwargs):
data = {key: finalizer(get(inds, bases), **kwargs)
for (key, finalizer, inds) in calls}
return xr.Dataset(data)
def test_get():
for tup in tuples:
get(1, tup)
def output_transform(output):
bbox_preds, cls_scores, centerness = get(
['bbox_pred', 'cls_score', 'centerness'], output, default=None)
return postprocess(bbox_preds, cls_scores, bbox_coder, centerness, topk=100,
iou_threshold=0.6, score_threshold=0.05, use_sigmoid=True)
def combine(base_tuples):
bases = tuple(np.stack(bs) for bs in zip(*base_tuples))
return tuple(f(*get(inds, bases)) for (f, inds) in calls)
def fn(a, p):
key = get("key", p, None)
seger = get("seger", p, lambda __: None)
return merge(a, {key: seger(item_code)})
lambda x: ("x" not in x) or ("y" not in x),
"XY plot does not have X and Y.",
),
(
{"histogram"},
lambda x: ("step" in x) and ("bins" in x),
"Histogram cannot have STEP and BINS.",
),
(
{"line", "scatter", "bar"},
lambda x: ("agg" in x["x"]) and ("agg" in x["y"]),
"XY plot cannot have an aggregation on X and Y.",
),
(
{"histogram", "pie"},
lambda x: ("agg" in get("x", x, {})) or ("agg" in get("y", x, {})) or
("agg" in get("axis", x, {})),
"Histograms and pie charts cannot have aggregations.",
),
(
{"histogram", "pie"},
lambda x: ("temporal" in get("x", x, {})) or
("temporal" in get("y", x, {})) or ("temporal" in get("axis", x, {})),
"Histograms and pie charts cannot have temporal axes.",
),
(
{"histogram"},
lambda x: ("x" in x) and ("y" in x),
"Histograms can have X or Y, not both.",
),
(
list(t.get(markov_features, markov_tables(tree), str(0.0))) +
list(t.get(markov_features, markov_tables(compress(tree)), str(0.0)))))
if __name__ == '__main__':
markov_features = ['markov_N', 'markov_R', 'markov_NR', 'markov_RN', 'markov_NN', 'markov_RR']
opts, args = getopt.getopt(sys.argv[1:], 'l:n', ['--language', '--normalize'])
language = None
normalize = False
for o,a in opts:
if o == '-l' or o == '--language':
language = a + ','
if o == '-n' or o == '--normalize':
normalize = True
# Generate header for feature values
print(('language,' if language is not None else '') +
(','.join(','.join(x[0] + (lambda s: '' if s[0] == '<' else '__' + s)(nf.__name__)
for nf in x[2]) for x in features) if normalize
else ','.join(x[0] for x in features)) + ',' +
','.join(markov_features) + ',' +
','.join([('compressed_' + f) for f in markov_features]))
# For each tree, compute its associated feature values
for line in fileinput.input(args):
tree = parse_sexp(line)[0]
print((language or '') + ','.join([str(','.join(str(float(f[1](tree)/float(nf(tree)))) for nf in f[2]) if normalize else f[1](tree)) for f in features] +
list(t.get(markov_features, markov_tables(tree), str(0.0))) +
list(t.get(markov_features, markov_tables(compress(tree)), str(0.0)))))
def make_price(item_code):
return compose(
excepts(StopIteration, first, lambda x: {}),
partial(get, item_code, default=[]),
lambda x: get(x, prices, {}),
)
def test_get():
for tup in tuples:
get(1, tup)
def main(us_hexagons, historical_sightings, model_file, debug, output_file):
logger.info(f"Reading hexagons from {us_hexagons.name}.")
squatchcast_locations = pd.read_csv(us_hexagons)
logger.info(f"Read {squatchcast_locations.shape[0]} hexagons.")
logger.info(
f"Reading historical sightings from {historical_sightings.name}.")
historical_sightings_frame = pd.read_csv(historical_sightings).query(
"~latitude.isnull()")
logger.info(
f"Read {historical_sightings_frame.shape[0]} historical_sightings.")
if debug:
logger.warning("Debug selected, pulling top five records.")
squatchcast_locations = squatchcast_locations.head()
num_locations = squatchcast_locations.shape[0]
lats = []
lons = []
logger.info("Extracting hexagon lat / lon values.")
for _, row in tqdm(squatchcast_locations.iterrows(), total=num_locations):
lat, lon = h3.h3_to_geo(row.hex_address)
lats.append(lat)
lons.append(lon)
squatchcast_locations.loc[:, "latitude"] = lats
squatchcast_locations.loc[:, "longitude"] = lons
session = requests.Session()
logger.info(f"Retrieving the weather for {num_locations} " "locations.")
weather_conditions = []
failed = 0
for _, row in tqdm(squatchcast_locations.iterrows(), total=num_locations):
request = create_weather_request(row.latitude, row.longitude,
DARK_SKY_KEY)
try:
weather_response = session.get(request)
# Make sure the response worked.
weather_response.raise_for_status()
# Now parse the json.
weather_conditions.append(weather_response.json())
except requests.HTTPError:
failed += 1
logger.info(f"{failed} requests to Dark Sky failed.")
# Extract the features a list of dicts. Plan is to turn that into a
# data frame and concatenate them to the squatchcast_locations.
logger.info("Unpacking weather results.")
squatchcast_features = []
for weather in tqdm(weather_conditions, total=num_locations):
# Append the current features.
daily = get_in(["daily", "data"], weather, [])
latitude = get("latitude", weather, np.nan)
longitude = get("longitude", weather, np.nan)
for conditions in daily:
get_condition = curry(get)(seq=conditions, default=np.nan)
squatchcast_features.append({
"date":
datetime.utcfromtimestamp(
get_condition("time")).strftime("%Y-%m-%d"),
"latitude":
latitude,
"longitude":
longitude,
"precip_type":
get("precipType", conditions, "no_precipitation"),
"temperature_high":
get_condition("temperatureHigh"),
"temperature_low":
get_condition("temperatureLow"),
"dew_point":
get_condition("dewPoint"),
"humidity":
get_condition("humidity"),
"cloud_cover":
get_condition("cloudCover"),
"moon_phase":
get_condition("moonPhase"),
"precip_intensity":
get_condition("precipIntensity"),
"precip_probability":
get_condition("precipProbability"),
"pressure":
get_condition("pressure"),
"uv_index":
get_condition("uvIndex"),
"visibility":
get_condition("visibility"),
"wind_bearing":
get_condition("windBearing"),
"wind_speed":
get_condition("windSpeed"),
})
squatchcast_frame = pd.DataFrame.from_records(squatchcast_features)
logger.info(f"Loading model from {model_file}.")
model = load(model_file)
logger.info(
f"Getting predictions for {squatchcast_frame.shape[0]} locations.")
with yaspin(text="👣 Calculating squatchcast. 👣", color="cyan"):
squatchcast_frame.loc[:, "squatchcast"] = model.predict_proba(
squatchcast_frame[RAW_FEATURES])[:, 1]
# Get the resoluton the US hexagon file is at and index the squatchcast
# results by that resolution.
us_resolution = h3.h3_get_resolution(
squatchcast_locations.head(1).hex_address[0])
squatchcast_frame.loc[:, "hex_address"] = np.apply_along_axis(
lambda x: h3.geo_to_h3(x[0], x[1], us_resolution),
axis=1,
arr=squatchcast_frame[["latitude", "longitude"]].values,
)
historical_sightings_frame.loc[:, "hex_address"] = np.apply_along_axis(
lambda x: h3.geo_to_h3(x[0], x[1], us_resolution),
axis=1,
arr=historical_sightings_frame[["latitude", "longitude"]].values,
)
historical_sightings_agg = (
historical_sightings_frame.groupby("hex_address").agg({
"number": "count"
}).reset_index())
# Now we need, for each day, a complete hexagonification of the US. We'll
# do this in a groupby and concatenate.
visualization_frames = []
for date, frame in squatchcast_frame.groupby("date"):
# Merge weather and US hexagons.
weather_location_merge = pd.merge(
squatchcast_locations.drop(columns=["latitude", "longitude"]),
frame,
on="hex_address",
how="left",
)
# Merge historical sightings.
visualization_frames.append(
pd.merge(
weather_location_merge,
historical_sightings_agg,
on="hex_address",
how="left",
).fillna(0).astype({
"number": "int"
}).rename(columns={"number": "historical_sightings"}))
pd.concat(visualization_frames).to_csv(output_file, index=False)
def get_embedded(sentences):
data = []
labels = []
for buffer in sentences:
if len(buffer) < 2:
continue
stack = []
stack.append(buffer.pop(0))
stack.append(buffer.pop(0))
while True:
concat = []
"""""" """""" """""" """""" """""" """""" """""" """""" """""" """""
Sw contains 18 elements
""" """""" """""" """""" """""" """""" """""" """""" """""" """""" ""
concat.append(
toolz.get(stack[-1].word, word2vec, np.zeros(300)
) if len(stack) > 0 else np.zeros(300))
concat.append(
toolz.get(stack[-2].word, word2vec, np.zeros(300)
) if len(stack) > 1 else np.zeros(300))
concat.append(
toolz.get(stack[-3].word, word2vec, np.zeros(300)
) if len(stack) > 2 else np.zeros(300))
concat.append(
toolz.get(buffer[0].word, word2vec, np.zeros(300)
) if len(buffer) > 0 else np.zeros(300))
concat.append(
toolz.get(buffer[1].word, word2vec, np.zeros(300)
) if len(buffer) > 1 else np.zeros(300))
concat.append(
toolz.get(buffer[2].word, word2vec, np.zeros(300)
) if len(buffer) > 2 else np.zeros(300))
concat.append(
toolz.get(stack[-1].children[0].
word, word2vec, np.zeros(300)) if len(stack) > 0
and len(stack[-1].children) > 0 else np.zeros(300))
concat.append(
toolz.get(stack[-1].children[1].
word, word2vec, np.zeros(300)) if len(stack) > 0
and len(stack[-1].children) > 1 else np.zeros(300))
concat.append(
toolz.get(stack[-1].children[-1].
word, word2vec, np.zeros(300)) if len(stack) > 0
and len(stack[-1].children) > 0 else np.zeros(300))
concat.append(
toolz.get(stack[-1].children[-2].
word, word2vec, np.zeros(300)) if len(stack) > 0
and len(stack[-1].children) > 1 else np.zeros(300))
concat.append(
toolz.get(stack[-2].children[0].
word, word2vec, np.zeros(300)) if len(stack) > 1
and len(stack[-2].children) > 0 else np.zeros(300))
concat.append(
toolz.get(stack[-2].children[1].
word, word2vec, np.zeros(300)) if len(stack) > 1
and len(stack[-2].children) > 1 else np.zeros(300))
concat.append(
toolz.get(stack[-2].children[-1].
word, word2vec, np.zeros(300)) if len(stack) > 1
and len(stack[-2].children) > 0 else np.zeros(300))
concat.append(
toolz.get(stack[-2].children[-2].
word, word2vec, np.zeros(300)) if len(stack) > 1
and len(stack[-2].children) > 1 else np.zeros(300))
concat.append(
toolz.get(stack[-1].children[0].children[0].
word, word2vec, np.zeros(300))
if len(stack) > 0 and len(stack[-1].children) > 0 and
len(stack[-1].children[0].children) > 0 else np.zeros(300))
concat.append(
toolz.get(stack[-1].children[-1].children[-1].
word, word2vec, np.zeros(300))
if len(stack) > 0 and len(stack[-1].children) > 0
and len(stack[-1].children[-1].children) > 0 else np.
zeros(300))
concat.append(
toolz.get(stack[-2].children[0].children[0].
word, word2vec, np.zeros(300))
if len(stack) > 1 and len(stack[-2].children) > 0 and
len(stack[-2].children[0].children) > 0 else np.zeros(300))
concat.append(
toolz.get(stack[-2].children[-1].children[-1].
word, word2vec, np.zeros(300))
if len(stack) > 1 and len(stack[-2].children) > 0
and len(stack[-2].children[-1].children) > 0 else np.
zeros(300))
"""""" """""" """""" """""" """""" """""" """""" """""" """""" """""
St contains 18 elements
""" """""" """""" """""" """""" """""" """""" """""" """""" """""" ""
concat.append(
toolz.get(stack[-1].tag, tag2vec, np.zeros(100)
) if len(stack) > 0 else np.zeros(100))
concat.append(
toolz.get(stack[-2].tag, tag2vec, np.zeros(100)
) if len(stack) > 1 else np.zeros(100))
concat.append(
toolz.get(stack[-3].tag, tag2vec, np.zeros(100)
) if len(stack) > 2 else np.zeros(100))
concat.append(
toolz.get(buffer[0].tag, tag2vec, np.zeros(100)
) if len(buffer) > 0 else np.zeros(100))
concat.append(
toolz.get(buffer[1].tag, tag2vec, np.zeros(100)
) if len(buffer) > 1 else np.zeros(100))
concat.append(
toolz.get(buffer[2].tag, tag2vec, np.zeros(100)
) if len(buffer) > 2 else np.zeros(100))
concat.append(
toolz.get(stack[-1].children[0].tag, tag2vec, np.zeros(100)
) if len(stack) > 0
and len(stack[-1].children) > 0 else np.zeros(100))
concat.append(
toolz.get(stack[-1].children[1].tag, tag2vec, np.zeros(100)
) if len(stack) > 0
and len(stack[-1].children) > 1 else np.zeros(100))
concat.append(
toolz.get(stack[-1].children[-1].tag, tag2vec, np.
zeros(100)) if len(stack) > 0
and len(stack[-1].children) > 0 else np.zeros(100))
concat.append(
toolz.get(stack[-1].children[-2].tag, tag2vec, np.
zeros(100)) if len(stack) > 0
and len(stack[-1].children) > 1 else np.zeros(100))
concat.append(
toolz.get(stack[-2].children[0].tag, tag2vec, np.zeros(100)
) if len(stack) > 1
and len(stack[-2].children) > 0 else np.zeros(100))
concat.append(
toolz.get(stack[-2].children[1].tag, tag2vec, np.zeros(100)
) if len(stack) > 1
and len(stack[-2].children) > 1 else np.zeros(100))
concat.append(
toolz.get(stack[-2].children[-1].tag, tag2vec, np.
zeros(100)) if len(stack) > 1
and len(stack[-2].children) > 0 else np.zeros(100))
concat.append(
toolz.get(stack[-2].children[-2].tag, tag2vec, np.
zeros(100)) if len(stack) > 1
and len(stack[-2].children) > 1 else np.zeros(100))
concat.append(
toolz.get(stack[-1].children[0].children[0].
tag, tag2vec, np.zeros(100))
if len(stack) > 0 and len(stack[-1].children) > 0 and
len(stack[-1].children[0].children) > 0 else np.zeros(100))
concat.append(
toolz.get(stack[-1].children[-1].children[-1].
tag, tag2vec, np.zeros(100))
if len(stack) > 0 and len(stack[-1].children) > 0
and len(stack[-1].children[-1].children) > 0 else np.
zeros(100))
concat.append(
toolz.get(stack[-2].children[0].children[0].
tag, tag2vec, np.zeros(100))
if len(stack) > 1 and len(stack[-2].children) > 0 and
len(stack[-2].children[0].children) > 0 else np.zeros(100))
concat.append(
toolz.get(stack[-2].children[-1].children[-1].
tag, tag2vec, np.zeros(100))
if len(stack) > 1 and len(stack[-2].children) > 0
and len(stack[-2].children[-1].children) > 0 else np.
zeros(100))
"""""" """""" """""" """""" """""" """""" """""" """""" """""" """""
Sl contains 12 elements
""" """""" """""" """""" """""" """""" """""" """""" """""" """""" ""
concat.append(
toolz.get(stack[-1].children[0].label, label2vec,
np.zeros(100)) if len(stack) > 0
and len(stack[-1].children) > 0 else np.zeros(100))
concat.append(
toolz.get(stack[-1].children[1].label, label2vec,
np.zeros(100)) if len(stack) > 0
and len(stack[-1].children) > 1 else np.zeros(100))
concat.append(
toolz.get(stack[-1].children[-1].label, label2vec,
np.zeros(100)) if len(stack) > 0
and len(stack[-1].children) > 0 else np.zeros(100))
concat.append(
toolz.get(stack[-1].children[-2].label, label2vec,
np.zeros(100)) if len(stack) > 0
and len(stack[-1].children) > 1 else np.zeros(100))
concat.append(
toolz.get(stack[-2].children[0].label, label2vec,
np.zeros(100)) if len(stack) > 1
and len(stack[-2].children) > 0 else np.zeros(100))
concat.append(
toolz.get(stack[-2].children[1].label, label2vec,
np.zeros(100)) if len(stack) > 1
and len(stack[-2].children) > 1 else np.zeros(100))
concat.append(
toolz.get(stack[-2].children[-1].label, label2vec,
np.zeros(100)) if len(stack) > 1
and len(stack[-2].children) > 0 else np.zeros(100))
concat.append(
toolz.get(stack[-2].children[-2].label, label2vec,
np.zeros(100)) if len(stack) > 1
and len(stack[-2].children) > 1 else np.zeros(100))
concat.append(
toolz.get(stack[-1].children[0].children[0].
label, label2vec, np.zeros(100))
if len(stack) > 0 and len(stack[-1].children) > 0 and
len(stack[-1].children[0].children) > 0 else np.zeros(100))
concat.append(
toolz.get(stack[-1].children[-1].children[-1].
label, label2vec, np.zeros(100))
if len(stack) > 0 and len(stack[-1].children) > 0
and len(stack[-1].children[-1].children) > 0 else np.
zeros(100))
concat.append(
toolz.get(stack[-2].children[0].children[0].
label, label2vec, np.zeros(100))
if len(stack) > 1 and len(stack[-2].children) > 0 and
len(stack[-2].children[0].children) > 0 else np.zeros(100))
concat.append(
toolz.get(stack[-2].children[-1].children[-1].
label, label2vec, np.zeros(100))
if len(stack) > 1 and len(stack[-2].children) > 0
and len(stack[-2].children[-1].children) > 0 else np.
zeros(100))
if len(stack) >= 2 and stack[-1].parent == stack[-2].index:
"""""" """""" """""" """""" """""" """""" """""" """""" """""" """""
RIGHT-ARC (stack[-2] => stack[-1])
""" """""" """""" """""" """""" """""" """""" """""" """""" """""" ""
data.append(np.concatenate(concat))
labels.append(label_id[stack[-1].label] +
len(label_id) * 0 + 1)
stack[-2].children.append(stack.pop(-1))
elif len(stack) >= 2 and stack[-2].parent == stack[-1].index:
"""""" """""" """""" """""" """""" """""" """""" """""" """""" """""
LEFT-ARC (stack[-2] <= stack[-1])
""" """""" """""" """""" """""" """""" """""" """""" """""" """""" ""
data.append(np.concatenate(concat))
labels.append(label_id[stack[-2].label] +
len(label_id) * 1 + 1)
stack[-1].children.append(stack.pop(-2))
else:
"""""" """""" """""" """""" """""" """""" """""" """""" """""" """""
SHIFT
""" """""" """""" """""" """""" """""" """""" """""" """""" """""" ""
if not buffer:
break
data.append(np.concatenate(concat))
labels.append(0)
stack.append(buffer.pop(0))
return np.array(data), np.array(labels)
def finalize(bases, **kwargs):
data = {
key: finalizer(get(inds, bases), **kwargs)
for (key, finalizer, inds) in calls
}
return xr.Dataset(data)
def fn(a, p):
key = get("key", p, None)
seger = get("seger", p, lambda __: None)
return merge(a, {key: seger(branch)})
def all_productions():
return {k: t.get(['id', 'name'], db[k]) for k in db}
def directeur(self) -> Profile | None:
direction = self.get_directeurs()
direction = [d for d in direction if d.is_directeur]
return toolz.get(0, direction, None)
def main(report_file, weather_file, weather_join_file):
weather_reader = csv.reader(weather_file)
# Load the weather into a dictionary.
weather_cache = {
# Extract the dict with the weather information.
(r[0], r[1]): get_in(["daily", "data", 0], json.loads(r[-1]), {})
for r in weather_reader
}
report_reader = csv.DictReader(report_file)
fieldnames = report_reader.fieldnames + [
"temperature_high", "temperature_mid", "temperature_low", "dew_point",
"humidity", "cloud_cover", "moon_phase", "precip_intensity",
"precip_probability", "precip_type", "pressure", "summary", "uv_index",
"visibility", "wind_bearing", "wind_speed"
]
writer = csv.DictWriter(weather_join_file, fieldnames=fieldnames)
writer.writeheader()
for line in report_reader:
weather = get((line["geohash"], line["date"]), weather_cache, {})
temperature_high = get("temperatureHigh", weather, None)
temperature_low = get("temperatureLow", weather, None)
line["temperature_high"] = temperature_high
line["temperature_mid"] = (
temperature_low + (temperature_high - temperature_low) /
2) if temperature_high and temperature_low else None
line["temperature_low"] = temperature_low
line["dew_point"] = get("dewPoint", weather, None)
line["humidity"] = get("humidity", weather, None)
line["cloud_cover"] = get("cloudCover", weather, None)
line["moon_phase"] = get("moonPhase", weather, None)
line["precip_intensity"] = get("precipIntensity", weather, None)
line["precip_probability"] = get("precipProbability", weather, None)
line["precip_type"] = get("precipType", weather, None)
line["pressure"] = get("pressure", weather, None)
line["summary"] = get("summary", weather, None)
line["uv_index"] = get("uvIndex", weather, None)
line["visibility"] = get("visibility", weather, None)
line["wind_bearing"] = get("windBearing", weather, None)
line["wind_speed"] = get("windSpeed", weather, None)
writer.writerow(line)
def featurize(tree, normalize=True):
markov_features = ['markov_N', 'markov_R', 'markov_NR', 'markov_RN', 'markov_NN', 'markov_RR']
tree = parse_sexp(tree)[0]
return (','.join([str(','.join(str(float(f[1](tree)/float(nf(tree)))) for nf in f[2]) if normalize else f[1](tree)) for f in features] +
list(t.get(markov_features, markov_tables(tree), str(0.0))) +
list(t.get(markov_features, markov_tables(compress(tree)), str(0.0)))))