This repository has been archived by the owner on Oct 31, 2023. It is now read-only.
-
Notifications
You must be signed in to change notification settings - Fork 23
/
viz_dataset.py
185 lines (148 loc) · 7.1 KB
/
viz_dataset.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
# Copyright (c) Facebook, Inc. and its affiliates.
import argparse
import os
import matplotlib
matplotlib.use("Agg") # noqa
import matplotlib.pyplot as plt
import matplotlib.patheffects as PathEffects
import numpy as np
from scipy.stats import gaussian_kde
import torch
import datasets
import toy_datasets
BBOXES = {
"citibike": (-74.03, -73.87, 40.65, 40.87),
"covid_nj_cases": (-75.60, -73.90, 38.90, 41.20),
"earthquakes_jp": (123.43, 149.18, 25.41, 45.98),
"pinwheel": (-4.0, 4.0, -4.0, 4.0),
"fmri": (0.0, 106.0, 0.0, 106.0),
}
MAPS = {
"citibike": "assets/manhattan_map.png",
"covid_nj_cases": "assets/nj_map.png",
"earthquakes_jp": "assets/jp_map.png",
"pinwheel": None,
"fmri": None,
"gmm": None,
}
FIGSIZE = 10
DPI = 300
def plot_coordinates(coords, S_std, S_mean, savepath, dataset_name):
coords = coords * S_std + S_mean
longs = coords[:, 0].detach().cpu().numpy()
lats = coords[:, 1].detach().cpu().numpy()
if MAPS[dataset_name]:
map_img = plt.imread(MAPS[dataset_name])
fig, ax = plt.subplots(figsize=(FIGSIZE, FIGSIZE * map_img.shape[0] / map_img.shape[1]))
ax.imshow(map_img, zorder=0, extent=BBOXES[dataset_name])
else:
fig, ax = plt.subplots(figsize=(FIGSIZE, FIGSIZE))
ax.scatter(longs, lats, s=1, alpha=0.4)
ax.set_xlim(BBOXES[dataset_name][0], BBOXES[dataset_name][1])
ax.set_ylim(BBOXES[dataset_name][2], BBOXES[dataset_name][3])
plt.axis('off')
os.makedirs(os.path.join(savepath, f"{dataset_name}"), exist_ok=True)
plt.savefig(os.path.join(savepath, f"{dataset_name}", f"{dataset_name}.png"), bbox_inches='tight', dpi=DPI)
plt.close()
def plot_kde(coords, S_std, S_mean, savepath, dataset_name, text=None, name=None):
name = f"{dataset_name}_density" if name is None else name
coords = coords * S_std.to(coords) + S_mean.to(coords)
longs = coords[:, 0].detach().cpu().numpy()
lats = coords[:, 1].detach().cpu().numpy()
if MAPS[dataset_name]:
map_img = plt.imread(MAPS[dataset_name])
fig, ax = plt.subplots(figsize=(FIGSIZE, FIGSIZE * map_img.shape[0] / map_img.shape[1]))
ax.imshow(map_img, zorder=0, extent=BBOXES[dataset_name])
else:
fig, ax = plt.subplots(figsize=(FIGSIZE, FIGSIZE))
kernel = gaussian_kde(np.stack([longs, lats], axis=0))
kernel.inv_cov = np.diag(np.diag(kernel.inv_cov))
X, Y = np.mgrid[BBOXES[dataset_name][0]:BBOXES[dataset_name][1]:100j, BBOXES[dataset_name][2]:BBOXES[dataset_name][3]:100j]
positions = np.vstack([X.ravel(), Y.ravel()])
Z = np.reshape(kernel(positions).T, X.shape)
ax.contourf(X, Y, Z, levels=10, alpha=0.6, cmap='RdGy')
ax.set_xlim(BBOXES[dataset_name][0], BBOXES[dataset_name][1])
ax.set_ylim(BBOXES[dataset_name][2], BBOXES[dataset_name][3])
if text is not None:
txt = ax.text(0.15, 0.9, text,
horizontalalignment="center",
verticalalignment="center",
transform=ax.transAxes,
size=16,
color='white')
txt.set_path_effects([PathEffects.withStroke(linewidth=5, foreground='black')])
plt.axis('off')
os.makedirs(os.path.join(savepath, f"{dataset_name}"), exist_ok=True)
plt.savefig(os.path.join(savepath, f"{dataset_name}", f"{name}.png"), bbox_inches='tight', dpi=DPI)
plt.close()
def plot_density(loglik_fn, spatial_locations, index, S_mean, S_std, savepath, dataset_name, device, text=None, fp64=False):
N = 50
x = np.linspace(BBOXES[dataset_name][0], BBOXES[dataset_name][1], N)
y = np.linspace(BBOXES[dataset_name][2], BBOXES[dataset_name][3], N)
s = np.stack([x, y], axis=1)
X, Y = np.meshgrid(s[:, 0], s[:, 1])
S = np.stack([X.reshape(-1), Y.reshape(-1)], axis=1)
S = torch.tensor(S).to(device)
S = S.double() if fp64 else S.float()
S = (S - S_mean.to(S)) / S_std.to(S)
logp = loglik_fn(S)
if MAPS[dataset_name]:
map_img = plt.imread(MAPS[dataset_name])
fig, ax = plt.subplots(figsize=(FIGSIZE, FIGSIZE * map_img.shape[0] / map_img.shape[1]))
ax.imshow(map_img, zorder=0, extent=BBOXES[dataset_name])
else:
fig, ax = plt.subplots(figsize=(FIGSIZE, FIGSIZE))
Z = logp.exp().detach().cpu().numpy().reshape(N, N)
ax.contourf(X, Y, Z, levels=20, alpha=0.7, cmap='RdGy')
spatial_locations = spatial_locations * np.array(S_std) + np.array(S_mean)
ax.scatter(spatial_locations[:, 0], spatial_locations[:, 1], s=20**2, alpha=1.0, marker="x", color="k")
ax.set_xlim(BBOXES[dataset_name][0], BBOXES[dataset_name][1])
ax.set_ylim(BBOXES[dataset_name][2], BBOXES[dataset_name][3])
if text:
txt = ax.text(0.15, 0.9, text,
horizontalalignment="center",
verticalalignment="center",
transform=ax.transAxes,
size=16,
color='white')
txt.set_path_effects([PathEffects.withStroke(linewidth=5, foreground='black')])
plt.axis('off')
os.makedirs(os.path.join(savepath, "figs"), exist_ok=True)
np.savez(f"{savepath}/figs/data{index}.npz", **{"X": X, "Y": Y, "Z": Z, "spatial_locations": spatial_locations})
plt.savefig(os.path.join(savepath, "figs", f"density{index}.png"), bbox_inches='tight', dpi=DPI)
plt.close()
def plot_intensities(list_of_event_times, list_of_intensities, list_of_timevals, savepath):
fig, axes = plt.subplots(nrows=len(list_of_event_times), figsize=(12, 1.5 * len(list_of_event_times)), sharex=True)
for ax, event_times, intensities, timevals in zip(axes, list_of_event_times, list_of_intensities, list_of_timevals):
ax.plot(timevals, intensities)
ax.vlines(event_times, ymin=0.0, ymax=100.0, linestyles="--", linewidth=1, alpha=0.35)
ax.set_xlim([timevals[0], timevals[-1]])
ax.set_ylim([0., np.max(intensities) + 0.2])
os.makedirs(os.path.dirname(savepath), exist_ok=True)
plt.savefig(savepath, bbox_inches='tight', dpi=DPI)
plt.close()
def load_data(data, split="train"):
if data == "citibike":
return datasets.Citibike(split=split)
elif data == "covid_nj_cases":
return datasets.CovidNJ(split=split)
elif data == "earthquakes_jp":
return datasets.Earthquakes(split=split)
elif data == "pinwheel":
return toy_datasets.PinwheelHawkes(split=split)
elif data == "gmm":
return toy_datasets.GMMHawkes(split=split)
elif data == "fmri":
return datasets.BOLD5000(split=split)
else:
raise ValueError(f"Unknown data option {data}")
if __name__ == "__main__":
parser = argparse.ArgumentParser()
parser.add_argument("--data", type=str, choices=MAPS.keys(), default="citibike")
args = parser.parse_args()
dataset = load_data(args.data)
savepath = "dataset_figs"
seq = dataset.__getitem__(0)
event_times, spatial_locations = seq[:, 0], seq[:, 1:]
plot_coordinates(spatial_locations, S_mean=dataset.S_mean, S_std=dataset.S_std, savepath=savepath, dataset_name=args.data)
plot_kde(spatial_locations, S_mean=dataset.S_mean, S_std=dataset.S_std, savepath=savepath, dataset_name=args.data)