-
Notifications
You must be signed in to change notification settings - Fork 0
/
analyze_linear.py
232 lines (198 loc) · 8.59 KB
/
analyze_linear.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
import nept
import numpy as np
from shapely.geometry import Point
import meta
import meta_session
from tasks import task
from utils import map_range
@task(infos=meta_session.all_infos, cache_saves="raw_linear")
def cache_raw_linear(info, *, position_byzone, lines):
"""Cache raw linearized position in .pkl"""
linear = {}
for trajectory in meta.trajectories:
full_line = lines[f"{trajectory}_with_feeders"]
feeder1 = f"{trajectory}_feeder1"
linear[feeder1] = position_byzone[feeder1].linearize(full_line)
linear[trajectory] = position_byzone[trajectory].linearize(full_line)
feeder2 = f"{trajectory}_feeder2"
linear[feeder2] = position_byzone[feeder2].linearize(full_line)
linear[f"{trajectory}_with_feeders"] = (
linear[feeder1].combine(linear[trajectory]).combine(linear[feeder2])
)
return linear
@task(infos=meta_session.all_infos, cache_saves="raw_matched_linear")
def cache_raw_matched_linear(info, *, raw_linear, lines):
u_start = lines["u"].project(Point(info.path_pts["shortcut1"]))
u_end = lines["u"].project(Point(info.path_pts["shortcut2"]))
u_dist = u_end - u_start
assert u_dist > 0, "u_end is past u_start"
full_shortcut_offset = lines["full_shortcut_feeder1"].length
full_shortcut_start = np.min(raw_linear["full_shortcut"].x) - full_shortcut_offset
full_shortcut_end = np.max(raw_linear["full_shortcut"].x) - full_shortcut_offset
full_shortcut_dist = full_shortcut_end - full_shortcut_start
assert full_shortcut_dist > 0, "full_shortcut_end is past full_shortcut_start"
# Check if we're in the weird case that shortening the full_shortcut would
# result in rebalancing past the original start/end
full_shortcut_barriers = [
lines["full_shortcut"].project(Point(info.path_pts["shortcut1"])),
lines["full_shortcut"].project(Point(info.path_pts["shortcut2"])),
]
full_shortcut_midpoint = sum(full_shortcut_barriers) / 2
if full_shortcut_dist > u_dist:
half_dist = u_dist / 2
if full_shortcut_midpoint - half_dist < full_shortcut_start:
dist = full_shortcut_barriers[0] - full_shortcut_start
assert full_shortcut_barriers[1] + dist < full_shortcut_end
full_shortcut_end = full_shortcut_barriers[1] + dist
elif full_shortcut_midpoint + half_dist > full_shortcut_end:
dist = full_shortcut_end - full_shortcut_barriers[1]
assert full_shortcut_barriers[0] - dist > full_shortcut_start
full_shortcut_start = full_shortcut_barriers[0] - dist
full_shortcut_dist = full_shortcut_end - full_shortcut_start
# Now we can rebalance as normal
if full_shortcut_dist > u_dist:
midpoint = (
lines["full_shortcut"].project(Point(info.path_pts["shortcut1"]))
+ lines["full_shortcut"].project(Point(info.path_pts["shortcut2"]))
) / 2
half_dist = u_dist / 2
full_shortcut_start = midpoint - half_dist
full_shortcut_end = midpoint + half_dist
else:
assert u_dist > full_shortcut_dist
midpoint = (u_end + u_start) / 2
half_dist = full_shortcut_dist / 2
u_start = midpoint - half_dist
u_end = midpoint + half_dist
# Add some extra points to deal with tuning curves falling off at the edges
binsize = (half_dist * 2) / 100
u_start -= meta.tc_extra_bins_before * binsize
u_end += meta.tc_extra_bins_after * binsize
full_shortcut_start -= meta.tc_extra_bins_before * binsize
full_shortcut_end += meta.tc_extra_bins_after * binsize
raw_u = raw_linear["u"]
matched_u = raw_u[(raw_u.x >= u_start) & (raw_u.x <= u_end)]
raw_full_shortcut = raw_linear["full_shortcut"]
raw_full_shortcut.x -= full_shortcut_offset
matched_full_shortcut = raw_full_shortcut[
(raw_full_shortcut.x >= full_shortcut_start)
& (raw_full_shortcut.x <= full_shortcut_end)
]
return {"u": matched_u, "full_shortcut": matched_full_shortcut}
@task(infos=meta_session.all_infos, cache_saves="tc_matched_linear")
def cache_tc_matched_linear(info, *, raw_matched_linear):
matched_linear = raw_matched_linear
for trajectory in matched_linear:
traj_linear = matched_linear[trajectory]
traj_linear.x[...] = map_range(
traj_linear.x,
from_min=np.min(traj_linear.x),
from_max=np.max(traj_linear.x),
to_min=meta.tc_linear_bin_edges[0],
to_max=meta.tc_linear_bin_edges[-1],
)
return matched_linear
@task(infos=meta_session.all_infos, cache_saves="matched_linear")
def cache_matched_linear(info, *, tc_matched_linear):
return {
trajectory: tc_matched_linear[trajectory][
(tc_matched_linear[trajectory].x >= meta.linear_bin_edges[0])
& (tc_matched_linear[trajectory].x <= meta.linear_bin_edges[-1])
]
for trajectory in meta.trajectories
}
@task(infos=meta_session.all_infos, cache_saves="joined_linear")
def cache_joined_linear(info, *, matched_linear):
joined_linear = nept.Position(
map_range(matched_linear["u"].x, from_min=0, from_max=100, to_min=0, to_max=50),
matched_linear["u"].time,
)
return joined_linear.combine(
nept.Position(
map_range(
matched_linear["full_shortcut"].x,
from_min=0,
from_max=100,
to_min=50,
to_max=100,
),
matched_linear["full_shortcut"].time,
)
)
def standardize_segment(info, x, out, line, start, stop):
if info.full_standard_maze:
standard_points = meta.full_standard_points
else:
standard_points = meta.short_standard_points
if start == "min":
from_min = np.amin(x)
else:
from_min = line.project(Point(info.path_pts[start]))
if stop == "max":
from_max = np.amax(x)
else:
from_max = line.project(Point(info.path_pts[stop]))
if start == "feeder1":
from_min += meta.feeder_dist
if stop == "feeder1":
from_max += meta.feeder_dist
if start == "feeder2":
from_min -= meta.feeder_dist
if stop == "feeder2":
from_max -= meta.feeder_dist
idx = (x >= from_min) & (x <= from_max)
out[idx] = map_range(
x[idx],
from_min=from_min,
from_max=from_max,
to_min=standard_points[start],
to_max=standard_points[stop],
)
@task(infos=meta_session.all_infos, cache_saves="tc_linear")
def cache_tc_linear(info, *, raw_linear, lines):
"""Cache standard linear in .pkl"""
linear = {
trajectory: raw_linear[f"{trajectory}_feeder1"]
.combine(raw_linear[trajectory])
.combine(raw_linear[f"{trajectory}_feeder2"])
for trajectory in meta.trajectories
}
full_line = lines["u_with_feeders"]
out = linear["u"].x
x = np.array(out)
standardize_segment(info, x, out, full_line, "min", "feeder1")
if info.full_standard_maze:
standardize_segment(info, x, out, full_line, "feeder1", "shortcut1")
standardize_segment(info, x, out, full_line, "shortcut1", "turn1")
else:
assert info.path_pts["feeder1"] == info.path_pts["shortcut1"]
standardize_segment(info, x, out, full_line, "feeder1", "turn1")
standardize_segment(info, x, out, full_line, "turn1", "turn2")
standardize_segment(info, x, out, full_line, "turn2", "shortcut2")
standardize_segment(info, x, out, full_line, "shortcut2", "feeder2")
standardize_segment(info, x, out, full_line, "feeder2", "max")
full_line = lines["full_shortcut_with_feeders"]
out = linear["full_shortcut"].x
x = np.array(out)
standardize_segment(info, x, out, full_line, "min", "feeder1")
if info.full_standard_maze:
standardize_segment(info, x, out, full_line, "feeder1", "shortcut1")
standardize_segment(info, x, out, full_line, "shortcut1", "shortcut2")
else:
standardize_segment(info, x, out, full_line, "feeder1", "shortcut2")
standardize_segment(info, x, out, full_line, "shortcut2", "feeder2")
standardize_segment(info, x, out, full_line, "feeder2", "max")
return linear
@task(infos=meta_session.all_infos, cache_saves="linear")
def cache_linear(info, *, tc_linear):
if info.full_standard_maze:
standard_points = meta.full_standard_points
else:
standard_points = meta.short_standard_points
return {
trajectory: tc_linear[trajectory][
(tc_linear[trajectory].x >= standard_points["feeder1"])
& (tc_linear[trajectory].x <= standard_points["feeder2"])
]
for trajectory in meta.trajectories
}