/
util.py
330 lines (294 loc) · 12.1 KB
/
util.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
import itertools
import numpy as np
import pandas as pd
import param
from ..core import Dataset
from ..core.boundingregion import BoundingBox
from ..core.data import PandasInterface, default_datatype
from ..core.operation import Operation
from ..core.sheetcoords import Slice
from ..core.util import (
cartesian_product,
datetime_types,
is_cyclic,
is_nan,
one_to_one,
sort_topologically,
)
def split_path(path):
"""
Split a Path type containing a single NaN separated path into
multiple subpaths.
"""
path = path.split(0, 1)[0]
values = path.dimension_values(0)
splits = np.concatenate([[0], np.where(np.isnan(values))[0]+1, [None]])
subpaths = []
data = PandasInterface.as_dframe(path)
for i in range(len(splits)-1):
end = splits[i+1]
slc = slice(splits[i], None if end is None else end-1)
subpath = data.iloc[slc]
if len(subpath):
subpaths.append(subpath)
return subpaths
def compute_slice_bounds(slices, scs, shape):
"""
Given a 2D selection consisting of slices/coordinates, a
SheetCoordinateSystem and the shape of the array returns a new
BoundingBox representing the sliced region.
"""
xidx, yidx = slices
ys, xs = shape
l, b, r, t = scs.bounds.lbrt()
xdensity, ydensity = scs.xdensity, scs.ydensity
xunit = (1./xdensity)
yunit = (1./ydensity)
if isinstance(l, datetime_types):
xunit = np.timedelta64(int(round(xunit)), scs._time_unit)
if isinstance(b, datetime_types):
yunit = np.timedelta64(int(round(yunit)), scs._time_unit)
if isinstance(xidx, slice):
l = l if xidx.start is None else max(l, xidx.start)
r = r if xidx.stop is None else min(r, xidx.stop)
if isinstance(yidx, slice):
b = b if yidx.start is None else max(b, yidx.start)
t = t if yidx.stop is None else min(t, yidx.stop)
bounds = BoundingBox(points=((l, b), (r, t)))
# Apply new bounds
slc = Slice(bounds, scs)
# Apply scalar and list indices
l, b, r, t = slc.compute_bounds(scs).lbrt()
if not isinstance(xidx, slice):
if not isinstance(xidx, (list, set)): xidx = [xidx]
if len(xidx) > 1:
xdensity = xdensity*(float(len(xidx))/xs)
ls, rs = [], []
for idx in xidx:
xc, _ = scs.closest_cell_center(idx, b)
ls.append(xc-xunit/2)
rs.append(xc+xunit/2)
l, r = np.min(ls), np.max(rs)
elif not isinstance(yidx, slice):
if not isinstance(yidx, (set, list)): yidx = [yidx]
if len(yidx) > 1:
ydensity = ydensity*(float(len(yidx))/ys)
bs, ts = [], []
for idx in yidx:
_, yc = scs.closest_cell_center(l, idx)
bs.append(yc-yunit/2)
ts.append(yc+yunit/2)
b, t = np.min(bs), np.max(ts)
return BoundingBox(points=((l, b), (r, t)))
def reduce_fn(x):
"""
Aggregation function to get the first non-zero value.
"""
values = x.values if isinstance(x, pd.Series) else x
for v in values:
if not is_nan(v):
return v
return np.nan
class categorical_aggregate2d(Operation):
"""
Generates a gridded Dataset of 2D aggregate arrays indexed by the
first two dimensions of the passed Element, turning all remaining
dimensions into value dimensions. The key dimensions of the
gridded array are treated as categorical indices. Useful for data
indexed by two independent categorical variables such as a table
of population values indexed by country and year. Data that is
indexed by continuous dimensions should be binned before
aggregation. The aggregation will retain the global sorting order
of both dimensions.
>> table = Table([('USA', 2000, 282.2), ('UK', 2005, 58.89)],
kdims=['Country', 'Year'], vdims=['Population'])
>> categorical_aggregate2d(table)
Dataset({'Country': ['USA', 'UK'], 'Year': [2000, 2005],
'Population': [[ 282.2 , np.nan], [np.nan, 58.89]]},
kdims=['Country', 'Year'], vdims=['Population'])
"""
datatype = param.List(default=['xarray', 'grid'], doc="""
The grid interface types to use when constructing the gridded Dataset.""")
@classmethod
def _get_coords(cls, obj):
"""
Get the coordinates of the 2D aggregate, maintaining the correct
sorting order.
"""
xdim, ydim = obj.dimensions(label=True)[:2]
xcoords = obj.dimension_values(xdim, False)
ycoords = obj.dimension_values(ydim, False)
if xcoords.dtype.kind not in 'SUO':
xcoords = np.sort(xcoords)
if ycoords.dtype.kind not in 'SUO':
return xcoords, np.sort(ycoords)
# Determine global orderings of y-values using topological sort
grouped = obj.groupby(xdim, container_type=dict,
group_type=Dataset).values()
orderings = {}
sort = True
for group in grouped:
vals = group.dimension_values(ydim, False)
if len(vals) == 1:
orderings[vals[0]] = [vals[0]]
else:
for i in range(len(vals)-1):
p1, p2 = vals[i:i+2]
orderings[p1] = [p2]
if sort:
if vals.dtype.kind in ('i', 'f'):
sort = (np.diff(vals)>=0).all()
else:
sort = np.array_equal(np.sort(vals), vals)
if sort or one_to_one(orderings, ycoords):
ycoords = np.sort(ycoords)
elif not is_cyclic(orderings):
coords = list(itertools.chain(*sort_topologically(orderings)))
ycoords = coords if len(coords) == len(ycoords) else np.sort(ycoords)
return np.asarray(xcoords), np.asarray(ycoords)
def _aggregate_dataset(self, obj):
"""
Generates a gridded Dataset from a column-based dataset and
lists of xcoords and ycoords
"""
xcoords, ycoords = self._get_coords(obj)
dim_labels = obj.dimensions(label=True)
vdims = obj.dimensions()[2:]
xdim, ydim = dim_labels[:2]
shape = (len(ycoords), len(xcoords))
nsamples = np.prod(shape)
grid_data = {xdim: xcoords, ydim: ycoords}
ys, xs = cartesian_product([ycoords, xcoords], copy=True)
data = {xdim: xs, ydim: ys}
for vdim in vdims:
values = np.empty(nsamples)
values[:] = np.nan
data[vdim.name] = values
dtype = default_datatype
dense_data = Dataset(data, kdims=obj.kdims, vdims=obj.vdims, datatype=[dtype])
concat_data = obj.interface.concatenate([dense_data, obj], datatype=dtype)
reindexed = concat_data.reindex([xdim, ydim], vdims)
if not reindexed:
agg = reindexed
df = PandasInterface.as_dframe(reindexed)
df = df.groupby([xdim, ydim], sort=False).first().reset_index()
agg = reindexed.clone(df)
# Convert data to a gridded dataset
for vdim in vdims:
grid_data[vdim.name] = agg.dimension_values(vdim).reshape(shape)
return agg.clone(grid_data, kdims=[xdim, ydim], vdims=vdims,
datatype=self.p.datatype)
def _aggregate_dataset_pandas(self, obj):
index_cols = [d.name for d in obj.kdims]
df = obj.data.set_index(index_cols).groupby(index_cols, sort=False).first()
label = 'unique' if len(df) == len(obj) else 'non-unique'
levels = self._get_coords(obj)
index = pd.MultiIndex.from_product(levels, names=df.index.names)
reindexed = df.reindex(index)
data = tuple(levels)
shape = tuple(d.shape[0] for d in data)
for vdim in obj.vdims:
data += (reindexed[vdim.name].values.reshape(shape).T,)
return obj.clone(data, datatype=self.p.datatype, label=label)
def _process(self, obj, key=None):
"""
Generates a categorical 2D aggregate by inserting NaNs at all
cross-product locations that do not already have a value assigned.
Returns a 2D gridded Dataset object.
"""
if isinstance(obj, Dataset) and obj.interface.gridded:
return obj
elif obj.ndims > 2:
raise ValueError("Cannot aggregate more than two dimensions")
elif len(obj.dimensions()) < 3:
raise ValueError("Must have at two dimensions to aggregate over"
"and one value dimension to aggregate on.")
obj = Dataset(obj, datatype=['dataframe'])
return self._aggregate_dataset_pandas(obj)
def circular_layout(nodes):
"""
Lay out nodes on a circle and add node index.
"""
N = len(nodes)
if not N:
return ([], [], [])
circ = np.pi/N*np.arange(N)*2
x = np.cos(circ)
y = np.sin(circ)
return (x, y, nodes)
def quadratic_bezier(start, end, c0=(0, 0), c1=(0, 0), steps=50):
"""
Compute quadratic bezier spline given start and end coordinate and
two control points.
"""
steps = np.linspace(0, 1, steps)
sx, sy = start
ex, ey = end
cx0, cy0 = c0
cx1, cy1 = c1
xs = ((1-steps)**3*sx + 3*((1-steps)**2)*steps*cx0 +
3*(1-steps)*steps**2*cx1 + steps**3*ex)
ys = ((1-steps)**3*sy + 3*((1-steps)**2)*steps*cy0 +
3*(1-steps)*steps**2*cy1 + steps**3*ey)
return np.column_stack([xs, ys])
def connect_edges_pd(graph):
"""
Given a Graph element containing abstract edges compute edge
segments directly connecting the source and target nodes. This
operation depends on pandas and is a lot faster than the pure
NumPy equivalent.
"""
edges = graph.dframe()
edges.index.name = 'graph_edge_index'
edges = edges.reset_index()
nodes = graph.nodes.dframe()
src, tgt = graph.kdims
x, y, idx = graph.nodes.kdims[:3]
df = pd.merge(edges, nodes, left_on=[src.name], right_on=[idx.name])
df = df.rename(columns={x.name: 'src_x', y.name: 'src_y'})
df = pd.merge(df, nodes, left_on=[tgt.name], right_on=[idx.name])
df = df.rename(columns={x.name: 'dst_x', y.name: 'dst_y'})
df = df.sort_values('graph_edge_index').drop(['graph_edge_index'], axis=1)
cols = ["src_x", "src_y", "dst_x", "dst_y"]
edge_segments = list(df[cols].values.reshape(df.index.size, 2, 2))
return edge_segments
def connect_tri_edges_pd(trimesh):
"""
Given a TriMesh element containing abstract edges compute edge
segments directly connecting the source and target nodes. This
operation depends on pandas and is a lot faster than the pure
NumPy equivalent.
"""
edges = trimesh.dframe().copy()
edges.index.name = 'trimesh_edge_index'
edges = edges.drop("color", errors="ignore", axis=1).reset_index()
nodes = trimesh.nodes.dframe().copy()
nodes.index.name = 'node_index'
nodes = nodes.drop(["color", "z"], errors="ignore", axis=1)
v1, v2, v3 = trimesh.kdims
x, y, idx = trimesh.nodes.kdims[:3]
df = pd.merge(edges, nodes, left_on=[v1.name], right_on=[idx.name])
df = df.rename(columns={x.name: 'x0', y.name: 'y0'})
df = pd.merge(df, nodes, left_on=[v2.name], right_on=[idx.name])
df = df.rename(columns={x.name: 'x1', y.name: 'y1'})
df = pd.merge(df, nodes, left_on=[v3.name], right_on=[idx.name])
df = df.rename(columns={x.name: 'x2', y.name: 'y2'})
df = df.sort_values('trimesh_edge_index').drop(['trimesh_edge_index'], axis=1)
return df[['x0', 'y0', 'x1', 'y1', 'x2', 'y2']]
def connect_edges(graph):
"""
Given a Graph element containing abstract edges compute edge
segments directly connecting the source and target nodes. This
operation just uses internal HoloViews operations and will be a
lot slower than the pandas equivalent.
"""
paths = []
for start, end in graph.array(graph.kdims):
start_ds = graph.nodes[:, :, start]
end_ds = graph.nodes[:, :, end]
if not len(start_ds) or not len(end_ds):
raise ValueError('Could not find node positions for all edges')
start = start_ds.array(start_ds.kdims[:2])
end = end_ds.array(end_ds.kdims[:2])
paths.append(np.array([start[0], end[0]]))
return paths