Spaces:
Runtime error
Runtime error
finished Alpha version
Browse filesThis view is limited to 50 files because it contains too many changes.
See raw diff
- .vscode/settings.json +3 -0
- Gradio_app.ipynb +230 -125
- app.py +226 -72
- data/.DS_Store +0 -0
- data/cached/BC_JARAX_2023-04-01T01:16:13.997267Z.mseed +0 -0
- data/cached/CE_24944_2023-04-01T01:16:14.044529Z.mseed +0 -0
- data/cached/CE_24945_2023-04-01T01:16:13.506381Z.mseed +0 -0
- data/cached/CI_ADO_2019-07-04T17:33:53.650962Z.mseed +0 -0
- data/cached/CI_ARV_2019-07-04T17:33:53.096986Z.mseed +0 -0
- data/cached/CI_CJV2_2023-04-01T01:16:16.011425Z.mseed +0 -0
- data/cached/CI_CWC_2019-07-04T17:33:47.189005Z.mseed +0 -0
- data/cached/CI_EDW2_2019-07-04T17:33:49.567241Z.mseed +0 -0
- data/cached/CI_FUR_2019-07-04T17:33:49.310305Z.mseed +0 -0
- data/cached/CI_GRA_2019-07-04T17:33:53.959922Z.mseed +0 -0
- data/cached/CI_GSC_2019-07-04T17:33:47.536363Z.mseed +0 -0
- data/cached/CI_HEC_2019-07-04T17:33:56.148977Z.mseed +0 -0
- data/cached/CI_ISA_2019-07-04T17:33:46.297658Z.mseed +0 -0
- data/cached/CI_JNH2_2023-04-01T01:16:13.664044Z.mseed +0 -0
- data/cached/CI_JRC2_2019-07-04T17:33:39.947494Z.mseed +0 -0
- data/cached/CI_LRL_2019-07-04T17:33:40.248999Z.mseed +0 -0
- data/cached/CI_LRR2_2023-04-01T01:16:15.095101Z.mseed +0 -0
- data/cached/CI_MPM_2019-07-04T17:33:40.443346Z.mseed +0 -0
- data/cached/CI_OSI_2019-07-04T17:33:57.203547Z.mseed +0 -0
- data/cached/CI_PUT_2023-04-01T01:16:16.558724Z.mseed +0 -0
- data/cached/CI_Q0013_2019-07-04T17:33:54.489098Z.mseed +0 -0
- data/cached/CI_Q0024_2019-07-04T17:33:55.620507Z.mseed +0 -0
- data/cached/CI_Q0035_2019-07-04T17:33:56.041452Z.mseed +0 -0
- data/cached/CI_Q0056_2019-07-04T17:33:50.407027Z.mseed +0 -0
- data/cached/CI_Q0061_2019-07-04T17:33:53.114695Z.mseed +0 -0
- data/cached/CI_Q0068_2019-07-04T17:33:46.411249Z.mseed +0 -0
- data/cached/CI_Q0072_2019-07-04T17:33:38.390421Z.mseed +0 -0
- data/cached/CI_RRX_2019-07-04T17:33:50.712219Z.mseed +0 -0
- data/cached/CI_SBB2_2023-04-01T01:16:15.609344Z.mseed +0 -0
- data/cached/CI_SHO_2019-07-04T17:33:51.673022Z.mseed +0 -0
- data/cached/CI_SLA_2019-07-04T17:33:40.190746Z.mseed +0 -0
- data/cached/CI_SRA_2023-04-01T01:16:14.887472Z.mseed +0 -0
- data/cached/CI_SRT_2019-07-04T17:33:38.029990Z.mseed +0 -0
- data/cached/CI_TIN_2019-07-04T17:33:55.946998Z.mseed +0 -0
- data/cached/CI_TOW2_2019-07-04T17:33:37.991033Z.mseed +0 -0
- data/cached/CI_VCS_2023-04-01T01:16:15.307897Z.mseed +0 -0
- data/cached/CI_VTV_2019-07-04T17:33:53.688913Z.mseed +0 -0
- data/cached/CI_WBM_2019-07-04T17:33:40.063644Z.mseed +0 -0
- data/cached/CI_WLS2_2023-04-01T01:16:13.623472Z.mseed +0 -0
- data/cached/CI_WMF_2019-07-04T17:33:41.867962Z.mseed +0 -0
- data/cached/CI_WRC2_2019-07-04T17:33:38.698107Z.mseed +0 -0
- data/cached/CI_WRV2_2019-07-04T17:33:40.843660Z.mseed +0 -0
- data/cached/CI_WVP2_2019-07-04T17:33:39.650418Z.mseed +0 -0
- data/cached/LB_DAC_2019-07-04T17:33:43.387184Z.mseed +0 -0
- data/cached/NN_FMT_2019-07-04T17:33:51.798341Z.mseed +0 -0
- data/cached/NN_GVN_2019-07-04T17:33:54.053591Z.mseed +0 -0
.vscode/settings.json
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"python.formatting.provider": "black"
|
3 |
+
}
|
Gradio_app.ipynb
CHANGED
@@ -2,42 +2,14 @@
|
|
2 |
"cells": [
|
3 |
{
|
4 |
"cell_type": "code",
|
5 |
-
"execution_count":
|
6 |
"metadata": {},
|
7 |
"outputs": [
|
8 |
-
{
|
9 |
-
"data": {
|
10 |
-
"text/plain": [
|
11 |
-
"'hi!'"
|
12 |
-
]
|
13 |
-
},
|
14 |
-
"execution_count": 1,
|
15 |
-
"metadata": {},
|
16 |
-
"output_type": "execute_result"
|
17 |
-
}
|
18 |
-
],
|
19 |
-
"source": [
|
20 |
-
"'hi!'"
|
21 |
-
]
|
22 |
-
},
|
23 |
-
{
|
24 |
-
"cell_type": "code",
|
25 |
-
"execution_count": 4,
|
26 |
-
"metadata": {},
|
27 |
-
"outputs": [
|
28 |
-
{
|
29 |
-
"name": "stderr",
|
30 |
-
"output_type": "stream",
|
31 |
-
"text": [
|
32 |
-
"/Users/anovosel/miniconda3/envs/phasehunter/lib/python3.11/site-packages/gradio/outputs.py:43: UserWarning: Usage of gradio.outputs is deprecated, and will not be supported in the future, please import your components from gradio.components\n",
|
33 |
-
" warnings.warn(\n"
|
34 |
-
]
|
35 |
-
},
|
36 |
{
|
37 |
"name": "stdout",
|
38 |
"output_type": "stream",
|
39 |
"text": [
|
40 |
-
"Running on local URL: http://127.0.0.1:
|
41 |
"\n",
|
42 |
"To create a public link, set `share=True` in `launch()`.\n"
|
43 |
]
|
@@ -45,7 +17,7 @@
|
|
45 |
{
|
46 |
"data": {
|
47 |
"text/html": [
|
48 |
-
"<div><iframe src=\"http://127.0.0.1:
|
49 |
],
|
50 |
"text/plain": [
|
51 |
"<IPython.core.display.HTML object>"
|
@@ -58,30 +30,9 @@
|
|
58 |
"data": {
|
59 |
"text/plain": []
|
60 |
},
|
61 |
-
"execution_count":
|
62 |
"metadata": {},
|
63 |
"output_type": "execute_result"
|
64 |
-
},
|
65 |
-
{
|
66 |
-
"name": "stdout",
|
67 |
-
"output_type": "stream",
|
68 |
-
"text": [
|
69 |
-
"Error in callback <function _draw_all_if_interactive at 0x1774d0ea0> (for post_execute):\n"
|
70 |
-
]
|
71 |
-
},
|
72 |
-
{
|
73 |
-
"ename": "KeyboardInterrupt",
|
74 |
-
"evalue": "",
|
75 |
-
"output_type": "error",
|
76 |
-
"traceback": [
|
77 |
-
"\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
|
78 |
-
"\u001b[0;31mKeyboardInterrupt\u001b[0m Traceback (most recent call last)",
|
79 |
-
"File \u001b[0;32m~/miniconda3/envs/phasehunter/lib/python3.11/site-packages/matplotlib/pyplot.py:120\u001b[0m, in \u001b[0;36m_draw_all_if_interactive\u001b[0;34m()\u001b[0m\n\u001b[1;32m 118\u001b[0m \u001b[39mdef\u001b[39;00m \u001b[39m_draw_all_if_interactive\u001b[39m():\n\u001b[1;32m 119\u001b[0m \u001b[39mif\u001b[39;00m matplotlib\u001b[39m.\u001b[39mis_interactive():\n\u001b[0;32m--> 120\u001b[0m draw_all()\n",
|
80 |
-
"File \u001b[0;32m~/miniconda3/envs/phasehunter/lib/python3.11/site-packages/matplotlib/_pylab_helpers.py:132\u001b[0m, in \u001b[0;36mGcf.draw_all\u001b[0;34m(cls, force)\u001b[0m\n\u001b[1;32m 130\u001b[0m \u001b[39mfor\u001b[39;00m manager \u001b[39min\u001b[39;00m \u001b[39mcls\u001b[39m\u001b[39m.\u001b[39mget_all_fig_managers():\n\u001b[1;32m 131\u001b[0m \u001b[39mif\u001b[39;00m force \u001b[39mor\u001b[39;00m manager\u001b[39m.\u001b[39mcanvas\u001b[39m.\u001b[39mfigure\u001b[39m.\u001b[39mstale:\n\u001b[0;32m--> 132\u001b[0m manager\u001b[39m.\u001b[39;49mcanvas\u001b[39m.\u001b[39;49mdraw_idle()\n",
|
81 |
-
"File \u001b[0;32m~/miniconda3/envs/phasehunter/lib/python3.11/site-packages/matplotlib/backend_bases.py:2082\u001b[0m, in \u001b[0;36mFigureCanvasBase.draw_idle\u001b[0;34m(self, *args, **kwargs)\u001b[0m\n\u001b[1;32m 2080\u001b[0m \u001b[39mif\u001b[39;00m \u001b[39mnot\u001b[39;00m \u001b[39mself\u001b[39m\u001b[39m.\u001b[39m_is_idle_drawing:\n\u001b[1;32m 2081\u001b[0m \u001b[39mwith\u001b[39;00m \u001b[39mself\u001b[39m\u001b[39m.\u001b[39m_idle_draw_cntx():\n\u001b[0;32m-> 2082\u001b[0m \u001b[39mself\u001b[39;49m\u001b[39m.\u001b[39;49mdraw(\u001b[39m*\u001b[39;49margs, \u001b[39m*\u001b[39;49m\u001b[39m*\u001b[39;49mkwargs)\n",
|
82 |
-
"File \u001b[0;32m~/miniconda3/envs/phasehunter/lib/python3.11/site-packages/matplotlib/backends/backend_agg.py:397\u001b[0m, in \u001b[0;36mFigureCanvasAgg.draw\u001b[0;34m(self)\u001b[0m\n\u001b[1;32m 395\u001b[0m \u001b[39mself\u001b[39m\u001b[39m.\u001b[39mrenderer\u001b[39m.\u001b[39mclear()\n\u001b[1;32m 396\u001b[0m \u001b[39m# Acquire a lock on the shared font cache.\u001b[39;00m\n\u001b[0;32m--> 397\u001b[0m \u001b[39mwith\u001b[39;49;00m RendererAgg\u001b[39m.\u001b[39;49mlock, \\\n\u001b[1;32m 398\u001b[0m (\u001b[39mself\u001b[39;49m\u001b[39m.\u001b[39;49mtoolbar\u001b[39m.\u001b[39;49m_wait_cursor_for_draw_cm() \u001b[39mif\u001b[39;49;00m \u001b[39mself\u001b[39;49m\u001b[39m.\u001b[39;49mtoolbar\n\u001b[1;32m 399\u001b[0m \u001b[39melse\u001b[39;49;00m nullcontext()):\n\u001b[1;32m 400\u001b[0m \u001b[39mself\u001b[39;49m\u001b[39m.\u001b[39;49mfigure\u001b[39m.\u001b[39;49mdraw(\u001b[39mself\u001b[39;49m\u001b[39m.\u001b[39;49mrenderer)\n\u001b[1;32m 401\u001b[0m \u001b[39m# A GUI class may be need to update a window using this draw, so\u001b[39;49;00m\n\u001b[1;32m 402\u001b[0m \u001b[39m# don't forget to call the superclass.\u001b[39;49;00m\n",
|
83 |
-
"\u001b[0;31mKeyboardInterrupt\u001b[0m: "
|
84 |
-
]
|
85 |
}
|
86 |
],
|
87 |
"source": [
|
@@ -95,6 +46,8 @@
|
|
95 |
"import torch\n",
|
96 |
"\n",
|
97 |
"from scipy.stats import gaussian_kde\n",
|
|
|
|
|
98 |
"\n",
|
99 |
"import obspy\n",
|
100 |
"from obspy.clients.fdsn import Client\n",
|
@@ -107,6 +60,7 @@
|
|
107 |
"\n",
|
108 |
"import matplotlib.pyplot as plt\n",
|
109 |
"import matplotlib.dates as mdates\n",
|
|
|
110 |
"\n",
|
111 |
"from glob import glob\n",
|
112 |
"\n",
|
@@ -123,21 +77,25 @@
|
|
123 |
"\n",
|
124 |
" return processed_input, p_phase, s_phase\n",
|
125 |
"\n",
|
126 |
-
"def mark_phases(waveform):\n",
|
|
|
|
|
|
|
|
|
127 |
" processed_input, p_phase, s_phase = make_prediction(waveform)\n",
|
128 |
"\n",
|
129 |
" # Create a plot of the waveform with the phases marked\n",
|
130 |
" if sum(processed_input[0][2] == 0): #if input is 1C\n",
|
131 |
" fig, ax = plt.subplots(nrows=2, figsize=(10, 2), sharex=True)\n",
|
132 |
"\n",
|
133 |
-
" ax[0].plot(processed_input[0][0])\n",
|
134 |
" ax[0].set_ylabel('Norm. Ampl.')\n",
|
135 |
"\n",
|
136 |
" else: #if input is 3C\n",
|
137 |
" fig, ax = plt.subplots(nrows=4, figsize=(10, 6), sharex=True)\n",
|
138 |
-
" ax[0].plot(processed_input[0][0])\n",
|
139 |
-
" ax[1].plot(processed_input[0][1])\n",
|
140 |
-
" ax[2].plot(processed_input[0][2])\n",
|
141 |
"\n",
|
142 |
" ax[0].set_ylabel('Z')\n",
|
143 |
" ax[1].set_ylabel('N')\n",
|
@@ -169,22 +127,39 @@
|
|
169 |
" plt.close(fig)\n",
|
170 |
" return image\n",
|
171 |
"\n",
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
172 |
"def variance_coefficient(residuals):\n",
|
173 |
" # calculate the variance of the residuals\n",
|
174 |
" var = residuals.var()\n",
|
175 |
-
" \n",
|
176 |
" # scale the variance to a coefficient between 0 and 1\n",
|
177 |
" coeff = 1 - (var / (residuals.max() - residuals.min()))\n",
|
178 |
-
" \n",
|
179 |
" return coeff\n",
|
180 |
"\n",
|
181 |
-
"def predict_on_section(client_name, timestamp, eq_lat, eq_lon, radius_km, source_depth_km, velocity_model):\n",
|
182 |
" distances, t0s, st_lats, st_lons, waveforms = [], [], [], [], []\n",
|
183 |
" \n",
|
184 |
" taup_model = TauPyModel(model=velocity_model)\n",
|
185 |
" client = Client(client_name)\n",
|
186 |
"\n",
|
187 |
" window = radius_km / 111.2\n",
|
|
|
188 |
"\n",
|
189 |
" assert eq_lat - window > -90 and eq_lat + window < 90, \"Latitude out of bounds\"\n",
|
190 |
" assert eq_lon - window > -180 and eq_lon + window < 180, \"Longitude out of bounds\"\n",
|
@@ -192,59 +167,107 @@
|
|
192 |
" starttime = obspy.UTCDateTime(timestamp)\n",
|
193 |
" endtime = starttime + 120\n",
|
194 |
"\n",
|
195 |
-
"
|
196 |
-
"
|
197 |
-
"
|
198 |
-
"
|
199 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
200 |
" \n",
|
201 |
" waveforms = []\n",
|
202 |
" cached_waveforms = glob(\"data/cached/*.mseed\")\n",
|
203 |
"\n",
|
204 |
" for network in inv:\n",
|
|
|
|
|
|
|
205 |
" for station in network:\n",
|
206 |
-
"
|
207 |
-
"
|
208 |
-
"\n",
|
209 |
-
" arrivals = taup_model.get_travel_times(source_depth_in_km=source_depth_km, \n",
|
210 |
-
" distance_in_degree=distance, \n",
|
211 |
-
" phase_list=[\"P\", \"S\"])\n",
|
212 |
"\n",
|
213 |
-
"
|
|
|
|
|
214 |
"\n",
|
215 |
-
"
|
216 |
-
" endtime = starttime + 60\n",
|
217 |
"\n",
|
|
|
|
|
|
|
218 |
" if f\"data/cached/{network.code}_{station.code}_{starttime}.mseed\" not in cached_waveforms:\n",
|
|
|
219 |
" waveform = client.get_waveforms(network=network.code, station=station.code, location=\"*\", channel=\"*\", \n",
|
220 |
" starttime=starttime, endtime=endtime)\n",
|
221 |
" waveform.write(f\"data/cached/{network.code}_{station.code}_{starttime}.mseed\", format=\"MSEED\")\n",
|
|
|
222 |
" else:\n",
|
|
|
223 |
" waveform = obspy.read(f\"data/cached/{network.code}_{station.code}_{starttime}.mseed\")\n",
|
224 |
-
"
|
225 |
-
"
|
226 |
-
"
|
227 |
-
"
|
228 |
-
" \n",
|
229 |
-
"
|
230 |
-
"
|
231 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
232 |
"\n",
|
233 |
-
" if len(waveform) == 3:\n",
|
234 |
-
" try:\n",
|
235 |
-
" waveform = prepare_waveform(np.stack([x.data for x in waveform]))\n",
|
236 |
-
" except:\n",
|
237 |
-
" continue\n",
|
238 |
-
" \n",
|
239 |
" distances.append(distance)\n",
|
240 |
" t0s.append(starttime)\n",
|
241 |
" st_lats.append(station.latitude)\n",
|
242 |
" st_lons.append(station.longitude)\n",
|
243 |
" waveforms.append(waveform)\n",
|
244 |
"\n",
|
245 |
-
"
|
246 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
247 |
"\n",
|
|
|
|
|
|
|
248 |
" with torch.no_grad():\n",
|
249 |
" waveforms_torch = torch.vstack(waveforms)\n",
|
250 |
" output = model(waveforms_torch)\n",
|
@@ -256,8 +279,31 @@
|
|
256 |
" p_max_confidence = np.min([p_phases[i::len(waveforms)].std() for i in range(len(waveforms))]) \n",
|
257 |
" s_max_confidence = np.min([s_phases[i::len(waveforms)].std() for i in range(len(waveforms))])\n",
|
258 |
"\n",
|
259 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
260 |
" for i in range(len(waveforms)):\n",
|
|
|
261 |
" current_P = p_phases[i::len(waveforms)]\n",
|
262 |
" current_S = s_phases[i::len(waveforms)]\n",
|
263 |
"\n",
|
@@ -275,17 +321,46 @@
|
|
275 |
" ax[0].set_ylabel('Z')\n",
|
276 |
"\n",
|
277 |
" ax[0].xaxis.set_major_formatter(mdates.DateFormatter('%H:%M:%S'))\n",
|
278 |
-
" ax[0].xaxis.set_major_locator(mdates.SecondLocator(interval=
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
279 |
" \n",
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
280 |
" ax[0].scatter(None, None, color='r', marker='|', label='P')\n",
|
281 |
" ax[0].scatter(None, None, color='b', marker='|', label='S')\n",
|
282 |
" ax[0].legend()\n",
|
283 |
"\n",
|
284 |
-
"
|
285 |
-
"
|
286 |
-
"
|
287 |
-
"
|
288 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
289 |
" fig.canvas.draw();\n",
|
290 |
" image = np.array(fig.canvas.renderer.buffer_rgba())\n",
|
291 |
" plt.close(fig)\n",
|
@@ -299,30 +374,54 @@
|
|
299 |
"model.eval()\n",
|
300 |
"\n",
|
301 |
"with gr.Blocks() as demo:\n",
|
302 |
-
" gr.
|
303 |
-
"
|
304 |
-
"
|
305 |
-
"
|
306 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
307 |
" \"\"\")\n",
|
308 |
-
" with gr.Tab(\"
|
309 |
-
"
|
310 |
-
"
|
311 |
-
"
|
312 |
-
"
|
313 |
-
"
|
314 |
-
"
|
315 |
-
"
|
316 |
-
"
|
317 |
-
"
|
|
|
|
|
|
|
318 |
"\n",
|
319 |
" button = gr.Button(\"Predict phases\")\n",
|
320 |
-
" outputs = gr.
|
321 |
" \n",
|
322 |
-
" button.click(mark_phases, inputs=inputs, outputs=outputs)\n",
|
323 |
" \n",
|
324 |
" with gr.Tab(\"Select earthquake from catalogue\"):\n",
|
325 |
-
" gr.Markdown(
|
|
|
326 |
" \n",
|
327 |
" client_inputs = gr.Dropdown(\n",
|
328 |
" choices = list(URL_MAPPINGS.keys()), \n",
|
@@ -331,6 +430,7 @@
|
|
331 |
" value = \"IRIS\",\n",
|
332 |
" interactive=True\n",
|
333 |
" )\n",
|
|
|
334 |
" with gr.Row(): \n",
|
335 |
"\n",
|
336 |
" timestamp_inputs = gr.Textbox(value='2019-07-04 17:33:49',\n",
|
@@ -364,13 +464,23 @@
|
|
364 |
" interactive=True)\n",
|
365 |
" \n",
|
366 |
" velocity_inputs = gr.Dropdown(\n",
|
367 |
-
" choices = ['1066a', '1066b', 'ak135',
|
|
|
|
|
368 |
" label=\"1D velocity model\", \n",
|
369 |
" info=\"Velocity model for station selection\",\n",
|
370 |
" value = \"1066a\",\n",
|
371 |
" interactive=True\n",
|
372 |
" )\n",
|
373 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
374 |
" \n",
|
375 |
" button = gr.Button(\"Predict phases\")\n",
|
376 |
" outputs_section = gr.Image(label='Waveforms with Phases Marked', type='numpy', interactive=False)\n",
|
@@ -378,15 +488,10 @@
|
|
378 |
" button.click(predict_on_section, \n",
|
379 |
" inputs=[client_inputs, timestamp_inputs, \n",
|
380 |
" eq_lat_inputs, eq_lon_inputs, \n",
|
381 |
-
" radius_inputs, source_depth_inputs,
|
|
|
382 |
" outputs=outputs_section)\n",
|
383 |
"\n",
|
384 |
-
" with gr.Tab(\"Predict on your own waveform\"):\n",
|
385 |
-
" gr.Markdown(\"\"\"\n",
|
386 |
-
" Please upload your waveform in .npy (numpy) format. \n",
|
387 |
-
" Your waveform should be sampled at 100 sps and have 3 (Z, N, E) or 1 (Z) channels.\n",
|
388 |
-
" \"\"\")\n",
|
389 |
-
"\n",
|
390 |
"demo.launch()"
|
391 |
]
|
392 |
},
|
|
|
2 |
"cells": [
|
3 |
{
|
4 |
"cell_type": "code",
|
5 |
+
"execution_count": 64,
|
6 |
"metadata": {},
|
7 |
"outputs": [
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
8 |
{
|
9 |
"name": "stdout",
|
10 |
"output_type": "stream",
|
11 |
"text": [
|
12 |
+
"Running on local URL: http://127.0.0.1:7904\n",
|
13 |
"\n",
|
14 |
"To create a public link, set `share=True` in `launch()`.\n"
|
15 |
]
|
|
|
17 |
{
|
18 |
"data": {
|
19 |
"text/html": [
|
20 |
+
"<div><iframe src=\"http://127.0.0.1:7904/\" width=\"100%\" height=\"500\" allow=\"autoplay; camera; microphone; clipboard-read; clipboard-write;\" frameborder=\"0\" allowfullscreen></iframe></div>"
|
21 |
],
|
22 |
"text/plain": [
|
23 |
"<IPython.core.display.HTML object>"
|
|
|
30 |
"data": {
|
31 |
"text/plain": []
|
32 |
},
|
33 |
+
"execution_count": 64,
|
34 |
"metadata": {},
|
35 |
"output_type": "execute_result"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
36 |
}
|
37 |
],
|
38 |
"source": [
|
|
|
46 |
"import torch\n",
|
47 |
"\n",
|
48 |
"from scipy.stats import gaussian_kde\n",
|
49 |
+
"from bmi_topography import Topography\n",
|
50 |
+
"import earthpy.spatial as es\n",
|
51 |
"\n",
|
52 |
"import obspy\n",
|
53 |
"from obspy.clients.fdsn import Client\n",
|
|
|
60 |
"\n",
|
61 |
"import matplotlib.pyplot as plt\n",
|
62 |
"import matplotlib.dates as mdates\n",
|
63 |
+
"from matplotlib.colors import LightSource\n",
|
64 |
"\n",
|
65 |
"from glob import glob\n",
|
66 |
"\n",
|
|
|
77 |
"\n",
|
78 |
" return processed_input, p_phase, s_phase\n",
|
79 |
"\n",
|
80 |
+
"def mark_phases(waveform, uploaded_file):\n",
|
81 |
+
"\n",
|
82 |
+
" if uploaded_file is not None:\n",
|
83 |
+
" waveform = uploaded_file.name\n",
|
84 |
+
"\n",
|
85 |
" processed_input, p_phase, s_phase = make_prediction(waveform)\n",
|
86 |
"\n",
|
87 |
" # Create a plot of the waveform with the phases marked\n",
|
88 |
" if sum(processed_input[0][2] == 0): #if input is 1C\n",
|
89 |
" fig, ax = plt.subplots(nrows=2, figsize=(10, 2), sharex=True)\n",
|
90 |
"\n",
|
91 |
+
" ax[0].plot(processed_input[0][0], color='black', lw=1)\n",
|
92 |
" ax[0].set_ylabel('Norm. Ampl.')\n",
|
93 |
"\n",
|
94 |
" else: #if input is 3C\n",
|
95 |
" fig, ax = plt.subplots(nrows=4, figsize=(10, 6), sharex=True)\n",
|
96 |
+
" ax[0].plot(processed_input[0][0], color='black', lw=1)\n",
|
97 |
+
" ax[1].plot(processed_input[0][1], color='black', lw=1)\n",
|
98 |
+
" ax[2].plot(processed_input[0][2], color='black', lw=1)\n",
|
99 |
"\n",
|
100 |
" ax[0].set_ylabel('Z')\n",
|
101 |
" ax[1].set_ylabel('N')\n",
|
|
|
127 |
" plt.close(fig)\n",
|
128 |
" return image\n",
|
129 |
"\n",
|
130 |
+
"def bin_distances(distances, bin_size=10):\n",
|
131 |
+
" # Bin the distances into groups of `bin_size` kilometers\n",
|
132 |
+
" binned_distances = {}\n",
|
133 |
+
" for i, distance in enumerate(distances):\n",
|
134 |
+
" bin_index = distance // bin_size\n",
|
135 |
+
" if bin_index not in binned_distances:\n",
|
136 |
+
" binned_distances[bin_index] = (distance, i)\n",
|
137 |
+
" elif i < binned_distances[bin_index][1]:\n",
|
138 |
+
" binned_distances[bin_index] = (distance, i)\n",
|
139 |
+
"\n",
|
140 |
+
" # Select the first distance in each bin and its index\n",
|
141 |
+
" first_distances = []\n",
|
142 |
+
" for bin_index in binned_distances:\n",
|
143 |
+
" first_distance, first_distance_index = binned_distances[bin_index]\n",
|
144 |
+
" first_distances.append(first_distance_index)\n",
|
145 |
+
" \n",
|
146 |
+
" return first_distances\n",
|
147 |
+
"\n",
|
148 |
"def variance_coefficient(residuals):\n",
|
149 |
" # calculate the variance of the residuals\n",
|
150 |
" var = residuals.var()\n",
|
|
|
151 |
" # scale the variance to a coefficient between 0 and 1\n",
|
152 |
" coeff = 1 - (var / (residuals.max() - residuals.min()))\n",
|
|
|
153 |
" return coeff\n",
|
154 |
"\n",
|
155 |
+
"def predict_on_section(client_name, timestamp, eq_lat, eq_lon, radius_km, source_depth_km, velocity_model, max_waveforms):\n",
|
156 |
" distances, t0s, st_lats, st_lons, waveforms = [], [], [], [], []\n",
|
157 |
" \n",
|
158 |
" taup_model = TauPyModel(model=velocity_model)\n",
|
159 |
" client = Client(client_name)\n",
|
160 |
"\n",
|
161 |
" window = radius_km / 111.2\n",
|
162 |
+
" max_waveforms = int(max_waveforms)\n",
|
163 |
"\n",
|
164 |
" assert eq_lat - window > -90 and eq_lat + window < 90, \"Latitude out of bounds\"\n",
|
165 |
" assert eq_lon - window > -180 and eq_lon + window < 180, \"Longitude out of bounds\"\n",
|
|
|
167 |
" starttime = obspy.UTCDateTime(timestamp)\n",
|
168 |
" endtime = starttime + 120\n",
|
169 |
"\n",
|
170 |
+
" try:\n",
|
171 |
+
" print('Starting to download inventory')\n",
|
172 |
+
" inv = client.get_stations(network=\"*\", station=\"*\", location=\"*\", channel=\"*H*\", \n",
|
173 |
+
" starttime=starttime, endtime=endtime, \n",
|
174 |
+
" minlatitude=(eq_lat-window), maxlatitude=(eq_lat+window),\n",
|
175 |
+
" minlongitude=(eq_lon-window), maxlongitude=(eq_lon+window), \n",
|
176 |
+
" level='station')\n",
|
177 |
+
" print('Finished downloading inventory')\n",
|
178 |
+
" except (IndexError, FDSNNoDataException, FDSNTimeoutException, FDSNInternalServerException):\n",
|
179 |
+
" fig, ax = plt.subplots()\n",
|
180 |
+
" ax.text(0.5,0.5,'Something is wrong with the data provider, try another')\n",
|
181 |
+
" fig.canvas.draw();\n",
|
182 |
+
" image = np.array(fig.canvas.renderer.buffer_rgba())\n",
|
183 |
+
" plt.close(fig)\n",
|
184 |
+
" return image\n",
|
185 |
" \n",
|
186 |
" waveforms = []\n",
|
187 |
" cached_waveforms = glob(\"data/cached/*.mseed\")\n",
|
188 |
"\n",
|
189 |
" for network in inv:\n",
|
190 |
+
" # Skip the SYntetic networks\n",
|
191 |
+
" if network.code == 'SY':\n",
|
192 |
+
" continue\n",
|
193 |
" for station in network:\n",
|
194 |
+
" print(f\"Processing {network.code}.{station.code}...\")\n",
|
195 |
+
" distance = locations2degrees(eq_lat, eq_lon, station.latitude, station.longitude)\n",
|
|
|
|
|
|
|
|
|
196 |
"\n",
|
197 |
+
" arrivals = taup_model.get_travel_times(source_depth_in_km=source_depth_km, \n",
|
198 |
+
" distance_in_degree=distance, \n",
|
199 |
+
" phase_list=[\"P\", \"S\"])\n",
|
200 |
"\n",
|
201 |
+
" if len(arrivals) > 0:\n",
|
|
|
202 |
"\n",
|
203 |
+
" starttime = obspy.UTCDateTime(timestamp) + arrivals[0].time - 15\n",
|
204 |
+
" endtime = starttime + 60\n",
|
205 |
+
" try:\n",
|
206 |
" if f\"data/cached/{network.code}_{station.code}_{starttime}.mseed\" not in cached_waveforms:\n",
|
207 |
+
" print('Downloading waveform')\n",
|
208 |
" waveform = client.get_waveforms(network=network.code, station=station.code, location=\"*\", channel=\"*\", \n",
|
209 |
" starttime=starttime, endtime=endtime)\n",
|
210 |
" waveform.write(f\"data/cached/{network.code}_{station.code}_{starttime}.mseed\", format=\"MSEED\")\n",
|
211 |
+
" print('Finished downloading and caching waveform')\n",
|
212 |
" else:\n",
|
213 |
+
" print('Reading cached waveform')\n",
|
214 |
" waveform = obspy.read(f\"data/cached/{network.code}_{station.code}_{starttime}.mseed\")\n",
|
215 |
+
" \n",
|
216 |
+
"\n",
|
217 |
+
" except (IndexError, FDSNNoDataException, FDSNTimeoutException, FDSNInternalServerException):\n",
|
218 |
+
" print(f'Skipping {network.code}_{station.code}_{starttime}')\n",
|
219 |
+
" continue\n",
|
220 |
+
" \n",
|
221 |
+
" waveform = waveform.select(channel=\"H[BH][ZNE]\")\n",
|
222 |
+
" waveform = waveform.merge(fill_value=0)\n",
|
223 |
+
" waveform = waveform[:3]\n",
|
224 |
+
" \n",
|
225 |
+
" len_check = [len(x.data) for x in waveform]\n",
|
226 |
+
" if len(set(len_check)) > 1:\n",
|
227 |
+
" continue\n",
|
228 |
+
"\n",
|
229 |
+
" if len(waveform) == 3:\n",
|
230 |
+
" try:\n",
|
231 |
+
" waveform = prepare_waveform(np.stack([x.data for x in waveform]))\n",
|
232 |
"\n",
|
|
|
|
|
|
|
|
|
|
|
|
|
233 |
" distances.append(distance)\n",
|
234 |
" t0s.append(starttime)\n",
|
235 |
" st_lats.append(station.latitude)\n",
|
236 |
" st_lons.append(station.longitude)\n",
|
237 |
" waveforms.append(waveform)\n",
|
238 |
"\n",
|
239 |
+
" print(f\"Added {network.code}.{station.code} to the list of waveforms\")\n",
|
240 |
+
"\n",
|
241 |
+
" except:\n",
|
242 |
+
" continue\n",
|
243 |
+
" \n",
|
244 |
+
" \n",
|
245 |
+
" # If there are no waveforms, return an empty plot\n",
|
246 |
+
" if len(waveforms) == 0:\n",
|
247 |
+
" fig, ax = plt.subplots()\n",
|
248 |
+
" ax.text(0.5,0.5,'No waveforms found')\n",
|
249 |
+
" fig.canvas.draw();\n",
|
250 |
+
" image = np.array(fig.canvas.renderer.buffer_rgba())\n",
|
251 |
+
" plt.close(fig)\n",
|
252 |
+
" return image\n",
|
253 |
+
" \n",
|
254 |
+
"\n",
|
255 |
+
" first_distances = bin_distances(distances, bin_size=10/111.2)\n",
|
256 |
+
"\n",
|
257 |
+
" # Edge case when there are way too many waveforms to process\n",
|
258 |
+
" selection_indexes = np.random.choice(first_distances, \n",
|
259 |
+
" np.min([len(first_distances), max_waveforms]),\n",
|
260 |
+
" replace=False)\n",
|
261 |
+
"\n",
|
262 |
+
" waveforms = np.array(waveforms)[selection_indexes]\n",
|
263 |
+
" distances = np.array(distances)[selection_indexes]\n",
|
264 |
+
" t0s = np.array(t0s)[selection_indexes]\n",
|
265 |
+
" st_lats = np.array(st_lats)[selection_indexes]\n",
|
266 |
+
" st_lons = np.array(st_lons)[selection_indexes]\n",
|
267 |
"\n",
|
268 |
+
" waveforms = [torch.tensor(waveform) for waveform in waveforms]\n",
|
269 |
+
"\n",
|
270 |
+
" print('Starting to run predictions')\n",
|
271 |
" with torch.no_grad():\n",
|
272 |
" waveforms_torch = torch.vstack(waveforms)\n",
|
273 |
" output = model(waveforms_torch)\n",
|
|
|
279 |
" p_max_confidence = np.min([p_phases[i::len(waveforms)].std() for i in range(len(waveforms))]) \n",
|
280 |
" s_max_confidence = np.min([s_phases[i::len(waveforms)].std() for i in range(len(waveforms))])\n",
|
281 |
"\n",
|
282 |
+
" print(f\"Starting plotting {len(waveforms)} waveforms\")\n",
|
283 |
+
" fig, ax = plt.subplots(nrows=1, ncols=3, figsize=(10, 3))\n",
|
284 |
+
"\n",
|
285 |
+
" # Plot topography\n",
|
286 |
+
" print('Fetching topography')\n",
|
287 |
+
" params = Topography.DEFAULT.copy()\n",
|
288 |
+
" extra_window = 0.5\n",
|
289 |
+
" params[\"south\"] = np.min([st_lats.min(), eq_lat])-extra_window\n",
|
290 |
+
" params[\"north\"] = np.max([st_lats.max(), eq_lat])+extra_window\n",
|
291 |
+
" params[\"west\"] = np.min([st_lons.min(), eq_lon])-extra_window\n",
|
292 |
+
" params[\"east\"] = np.max([st_lons.max(), eq_lon])+extra_window\n",
|
293 |
+
"\n",
|
294 |
+
" topo_map = Topography(**params)\n",
|
295 |
+
" topo_map.fetch()\n",
|
296 |
+
" topo_map.load()\n",
|
297 |
+
"\n",
|
298 |
+
" print('Plotting topo')\n",
|
299 |
+
" hillshade = es.hillshade(topo_map.da[0], altitude=10)\n",
|
300 |
+
" \n",
|
301 |
+
" topo_map.da.plot(ax = ax[1], cmap='Greys', add_colorbar=False, add_labels=False)\n",
|
302 |
+
" topo_map.da.plot(ax = ax[2], cmap='Greys', add_colorbar=False, add_labels=False)\n",
|
303 |
+
" ax[1].imshow(hillshade, cmap=\"Greys\", alpha=0.5)\n",
|
304 |
+
"\n",
|
305 |
" for i in range(len(waveforms)):\n",
|
306 |
+
" print(f\"Plotting waveform {i+1}/{len(waveforms)}\")\n",
|
307 |
" current_P = p_phases[i::len(waveforms)]\n",
|
308 |
" current_S = s_phases[i::len(waveforms)]\n",
|
309 |
"\n",
|
|
|
321 |
" ax[0].set_ylabel('Z')\n",
|
322 |
"\n",
|
323 |
" ax[0].xaxis.set_major_formatter(mdates.DateFormatter('%H:%M:%S'))\n",
|
324 |
+
" ax[0].xaxis.set_major_locator(mdates.SecondLocator(interval=20))\n",
|
325 |
+
"\n",
|
326 |
+
" delta_t = t0s[i].timestamp - obspy.UTCDateTime(timestamp).timestamp\n",
|
327 |
+
"\n",
|
328 |
+
" velocity_p = (distances[i]*111.2)/(delta_t+current_P.mean()*60).item()\n",
|
329 |
+
" velocity_s = (distances[i]*111.2)/(delta_t+current_S.mean()*60).item()\n",
|
330 |
+
"\n",
|
331 |
+
" print(f\"Station {st_lats[i]}, {st_lons[i]} has P velocity {velocity_p} and S velocity {velocity_s}\")\n",
|
332 |
" \n",
|
333 |
+
" # Generate an array from st_lat to eq_lat and from st_lon to eq_lon\n",
|
334 |
+
" x = np.linspace(st_lons[i], eq_lon, 50)\n",
|
335 |
+
" y = np.linspace(st_lats[i], eq_lat, 50)\n",
|
336 |
+
" \n",
|
337 |
+
" # Plot the array\n",
|
338 |
+
" ax[1].scatter(x, y, c=np.zeros_like(x)+velocity_p, alpha=0.5, vmin=0, vmax=8)\n",
|
339 |
+
" ax[2].scatter(x, y, c=np.zeros_like(x)+velocity_s, alpha=0.5, vmin=0, vmax=8)\n",
|
340 |
+
"\n",
|
341 |
+
" # Add legend\n",
|
342 |
" ax[0].scatter(None, None, color='r', marker='|', label='P')\n",
|
343 |
" ax[0].scatter(None, None, color='b', marker='|', label='S')\n",
|
344 |
" ax[0].legend()\n",
|
345 |
"\n",
|
346 |
+
" print('Plotting stations')\n",
|
347 |
+
" for i in range(1,3):\n",
|
348 |
+
" ax[i].scatter(st_lons, st_lats, color='b', label='Stations')\n",
|
349 |
+
" ax[i].scatter(eq_lon, eq_lat, color='r', marker='*', label='Earthquake')\n",
|
350 |
+
"\n",
|
351 |
+
" # Generate colorbar for the velocity plot\n",
|
352 |
+
" cbar = plt.colorbar(ax[1].scatter(None, None, c=velocity_p, alpha=0.5, vmin=0, vmax=8), ax=ax[1])\n",
|
353 |
+
" cbar.set_label('P Velocity (km/s)')\n",
|
354 |
+
" ax[1].set_title('P Velocity')\n",
|
355 |
+
"\n",
|
356 |
+
" cbar = plt.colorbar(ax[2].scatter(None, None, c=velocity_s, alpha=0.5, vmin=0, vmax=8), ax=ax[2])\n",
|
357 |
+
" cbar.set_label('S Velocity (km/s)')\n",
|
358 |
+
" ax[2].set_title('S Velocity')\n",
|
359 |
+
"\n",
|
360 |
+
"\n",
|
361 |
+
"\n",
|
362 |
+
" plt.subplots_adjust(hspace=0., wspace=0.5)\n",
|
363 |
+
"\n",
|
364 |
" fig.canvas.draw();\n",
|
365 |
" image = np.array(fig.canvas.renderer.buffer_rgba())\n",
|
366 |
" plt.close(fig)\n",
|
|
|
374 |
"model.eval()\n",
|
375 |
"\n",
|
376 |
"with gr.Blocks() as demo:\n",
|
377 |
+
" gr.HTML(\"\"\"<h1>PhaseHunter</h1>\n",
|
378 |
+
"<p>This app allows one to detect <mark style=\"background-color: red; color: white;\">P</mark> and <mark style=\"background-color: blue; color: white;\">S</mark> seismic phases along with \n",
|
379 |
+
"\n",
|
380 |
+
"<span style=\"font-size: 24px; font-weight: bold;\">\n",
|
381 |
+
" <span style=\"color: #4c4c4c; text-shadow: 1px 1px 0 #ccc, -1px -1px 0 #ccc, 1px -1px 0 #ccc, -1px 1px 0 #ccc;\">u</span>\n",
|
382 |
+
" <span style=\"color: #808080; text-shadow: 1px 1px 0 #bbb, -1px -1px 0 #bbb, 1px -1px 0 #bbb, -1px 1px 0 #bbb;\">n</span>\n",
|
383 |
+
" <span style=\"color: #b3b3b3; text-shadow: 1px 1px 0 #aaa, -1px -1px 0 #aaa, 1px -1px 0 #aaa, -1px 1px 0 #aaa;\">c</span>\n",
|
384 |
+
" <span style=\"color: #d9d9d9; text-shadow: 1px 1px 0 #ccc, -1px -1px 0 #ccc, 1px -1px 0 #ccc, -1px 1px 0 #ccc;\">e</span>\n",
|
385 |
+
" <span style=\"color: #f2f2f2; text-shadow: 1px 1px 0 #eee, -1px -1px 0 #eee, 1px -1px 0 #eee, -1px 1px 0 #eee;\">r</span>\n",
|
386 |
+
" <span style=\"color: #d9d9d9; text-shadow: 1px 1px 0 #ccc, -1px -1px 0 #ccc, 1px -1px 0 #ccc, -1px 1px 0 #ccc;\">t</span>\n",
|
387 |
+
" <span style=\"color: #b3b3b3; text-shadow: 1px 1px 0 #aaa, -1px -1px 0 #aaa, 1px -1px 0 #aaa, -1px 1px 0 #aaa;\">a</span>\n",
|
388 |
+
" <span style=\"color: #808080; text-shadow: 1px 1px 0 #bbb, -1px -1px 0 #bbb, 1px -1px 0 #bbb, -1px 1px 0 #bbb;\">i</span>\n",
|
389 |
+
" <span style=\"color: #4c4c4c; text-shadow: 1px 1px 0 #ccc, -1px -1px 0 #ccc, 1px -1px 0 #ccc, -1px 1px 0 #ccc;\">n</span>\n",
|
390 |
+
" <span style=\"color: #808080; text-shadow: 1px 1px 0 #bbb, -1px -1px 0 #bbb, 1px -1px 0 #bbb, -1px 1px 0 #bbb;\">t</span>\n",
|
391 |
+
" <span style=\"color: #b3b3b3; text-shadow: 1px 1px 0 #aaa, -1px -1px 0 #aaa, 1px -1px 0 #aaa, -1px 1px 0 #aaa;\">y</span>\n",
|
392 |
+
" \n",
|
393 |
+
"</span>\n",
|
394 |
+
" of the detection.</p>\n",
|
395 |
+
"<ol>\n",
|
396 |
+
" <li>By selecting one of the sample waveforms.</li>\n",
|
397 |
+
" <li>By uploading your own waveform.</li>\n",
|
398 |
+
" <li>By selecting an earthquake from the global earthquake catalogue.</li>\n",
|
399 |
+
"</ol>\n",
|
400 |
+
"<p>Please upload your waveform in <code>.npy</code> (numpy) format.</p>\n",
|
401 |
+
"<p>Your waveform should be sampled at 100 samples per second and have 3 (Z, N, E) or 1 (Z) channels. If your file is longer than 60 seconds, the app will only use the first 60 seconds of the waveform.</p>\n",
|
402 |
" \"\"\")\n",
|
403 |
+
" with gr.Tab(\"Try on a single station\"):\n",
|
404 |
+
" with gr.Row(): \n",
|
405 |
+
" # Define the input and output types for Gradio\n",
|
406 |
+
" inputs = gr.Dropdown(\n",
|
407 |
+
" [\"data/sample/sample_0.npy\", \n",
|
408 |
+
" \"data/sample/sample_1.npy\", \n",
|
409 |
+
" \"data/sample/sample_2.npy\"], \n",
|
410 |
+
" label=\"Sample waveform\", \n",
|
411 |
+
" info=\"Select one of the samples\",\n",
|
412 |
+
" value = \"data/sample/sample_0.npy\"\n",
|
413 |
+
" )\n",
|
414 |
+
"\n",
|
415 |
+
" upload = gr.File(label=\"Or upload your own waveform\")\n",
|
416 |
"\n",
|
417 |
" button = gr.Button(\"Predict phases\")\n",
|
418 |
+
" outputs = gr.Image(label='Waveform with Phases Marked', type='numpy', interactive=False)\n",
|
419 |
" \n",
|
420 |
+
" button.click(mark_phases, inputs=[inputs, upload], outputs=outputs)\n",
|
421 |
" \n",
|
422 |
" with gr.Tab(\"Select earthquake from catalogue\"):\n",
|
423 |
+
" gr.Markdown(\"\"\"Select an earthquake from the global earthquake catalogue and the app will download the waveform from the FDSN client of your choice.\n",
|
424 |
+
" \"\"\")\n",
|
425 |
" \n",
|
426 |
" client_inputs = gr.Dropdown(\n",
|
427 |
" choices = list(URL_MAPPINGS.keys()), \n",
|
|
|
430 |
" value = \"IRIS\",\n",
|
431 |
" interactive=True\n",
|
432 |
" )\n",
|
433 |
+
"\n",
|
434 |
" with gr.Row(): \n",
|
435 |
"\n",
|
436 |
" timestamp_inputs = gr.Textbox(value='2019-07-04 17:33:49',\n",
|
|
|
464 |
" interactive=True)\n",
|
465 |
" \n",
|
466 |
" velocity_inputs = gr.Dropdown(\n",
|
467 |
+
" choices = ['1066a', '1066b', 'ak135', \n",
|
468 |
+
" 'ak135f', 'herrin', 'iasp91', \n",
|
469 |
+
" 'jb', 'prem', 'pwdk'], \n",
|
470 |
" label=\"1D velocity model\", \n",
|
471 |
" info=\"Velocity model for station selection\",\n",
|
472 |
" value = \"1066a\",\n",
|
473 |
" interactive=True\n",
|
474 |
" )\n",
|
475 |
+
"\n",
|
476 |
+
" max_waveforms_inputs = gr.Slider(minimum=1,\n",
|
477 |
+
" maximum=100,\n",
|
478 |
+
" value=10,\n",
|
479 |
+
" label=\"Max waveforms per section\",\n",
|
480 |
+
" step=1,\n",
|
481 |
+
" info=\"Maximum number of waveforms to show per section\\n (to avoid long prediction times)\",\n",
|
482 |
+
" interactive=True,\n",
|
483 |
+
" )\n",
|
484 |
" \n",
|
485 |
" button = gr.Button(\"Predict phases\")\n",
|
486 |
" outputs_section = gr.Image(label='Waveforms with Phases Marked', type='numpy', interactive=False)\n",
|
|
|
488 |
" button.click(predict_on_section, \n",
|
489 |
" inputs=[client_inputs, timestamp_inputs, \n",
|
490 |
" eq_lat_inputs, eq_lon_inputs, \n",
|
491 |
+
" radius_inputs, source_depth_inputs, \n",
|
492 |
+
" velocity_inputs, max_waveforms_inputs],\n",
|
493 |
" outputs=outputs_section)\n",
|
494 |
"\n",
|
|
|
|
|
|
|
|
|
|
|
|
|
495 |
"demo.launch()"
|
496 |
]
|
497 |
},
|
app.py
CHANGED
@@ -8,6 +8,8 @@ from phasehunter.data_preparation import prepare_waveform
|
|
8 |
import torch
|
9 |
|
10 |
from scipy.stats import gaussian_kde
|
|
|
|
|
11 |
|
12 |
import obspy
|
13 |
from obspy.clients.fdsn import Client
|
@@ -20,6 +22,7 @@ from obspy.clients.fdsn.header import URL_MAPPINGS
|
|
20 |
|
21 |
import matplotlib.pyplot as plt
|
22 |
import matplotlib.dates as mdates
|
|
|
23 |
|
24 |
from glob import glob
|
25 |
|
@@ -36,21 +39,25 @@ def make_prediction(waveform):
|
|
36 |
|
37 |
return processed_input, p_phase, s_phase
|
38 |
|
39 |
-
def mark_phases(waveform):
|
|
|
|
|
|
|
|
|
40 |
processed_input, p_phase, s_phase = make_prediction(waveform)
|
41 |
|
42 |
# Create a plot of the waveform with the phases marked
|
43 |
if sum(processed_input[0][2] == 0): #if input is 1C
|
44 |
fig, ax = plt.subplots(nrows=2, figsize=(10, 2), sharex=True)
|
45 |
|
46 |
-
ax[0].plot(processed_input[0][0])
|
47 |
ax[0].set_ylabel('Norm. Ampl.')
|
48 |
|
49 |
else: #if input is 3C
|
50 |
fig, ax = plt.subplots(nrows=4, figsize=(10, 6), sharex=True)
|
51 |
-
ax[0].plot(processed_input[0][0])
|
52 |
-
ax[1].plot(processed_input[0][1])
|
53 |
-
ax[2].plot(processed_input[0][2])
|
54 |
|
55 |
ax[0].set_ylabel('Z')
|
56 |
ax[1].set_ylabel('N')
|
@@ -82,22 +89,39 @@ def mark_phases(waveform):
|
|
82 |
plt.close(fig)
|
83 |
return image
|
84 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
85 |
def variance_coefficient(residuals):
|
86 |
# calculate the variance of the residuals
|
87 |
var = residuals.var()
|
88 |
-
|
89 |
# scale the variance to a coefficient between 0 and 1
|
90 |
coeff = 1 - (var / (residuals.max() - residuals.min()))
|
91 |
-
|
92 |
return coeff
|
93 |
|
94 |
-
def predict_on_section(client_name, timestamp, eq_lat, eq_lon, radius_km, source_depth_km, velocity_model):
|
95 |
distances, t0s, st_lats, st_lons, waveforms = [], [], [], [], []
|
96 |
|
97 |
taup_model = TauPyModel(model=velocity_model)
|
98 |
client = Client(client_name)
|
99 |
|
100 |
window = radius_km / 111.2
|
|
|
101 |
|
102 |
assert eq_lat - window > -90 and eq_lat + window < 90, "Latitude out of bounds"
|
103 |
assert eq_lon - window > -180 and eq_lon + window < 180, "Longitude out of bounds"
|
@@ -105,59 +129,107 @@ def predict_on_section(client_name, timestamp, eq_lat, eq_lon, radius_km, source
|
|
105 |
starttime = obspy.UTCDateTime(timestamp)
|
106 |
endtime = starttime + 120
|
107 |
|
108 |
-
|
109 |
-
|
110 |
-
|
111 |
-
|
112 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
113 |
|
114 |
waveforms = []
|
115 |
cached_waveforms = glob("data/cached/*.mseed")
|
116 |
|
117 |
for network in inv:
|
|
|
|
|
|
|
118 |
for station in network:
|
119 |
-
|
120 |
-
|
121 |
|
122 |
-
|
123 |
-
|
124 |
-
|
125 |
|
126 |
-
|
127 |
-
|
128 |
-
starttime = obspy.UTCDateTime(timestamp) + arrivals[0].time - 15
|
129 |
-
endtime = starttime + 60
|
130 |
|
|
|
|
|
|
|
131 |
if f"data/cached/{network.code}_{station.code}_{starttime}.mseed" not in cached_waveforms:
|
|
|
132 |
waveform = client.get_waveforms(network=network.code, station=station.code, location="*", channel="*",
|
133 |
starttime=starttime, endtime=endtime)
|
134 |
waveform.write(f"data/cached/{network.code}_{station.code}_{starttime}.mseed", format="MSEED")
|
|
|
135 |
else:
|
|
|
136 |
waveform = obspy.read(f"data/cached/{network.code}_{station.code}_{starttime}.mseed")
|
137 |
-
|
138 |
-
|
139 |
-
|
140 |
-
|
141 |
-
|
142 |
-
|
143 |
-
|
144 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
145 |
|
146 |
-
if len(waveform) == 3:
|
147 |
-
try:
|
148 |
-
waveform = prepare_waveform(np.stack([x.data for x in waveform]))
|
149 |
-
except:
|
150 |
-
continue
|
151 |
-
|
152 |
distances.append(distance)
|
153 |
t0s.append(starttime)
|
154 |
st_lats.append(station.latitude)
|
155 |
st_lons.append(station.longitude)
|
156 |
waveforms.append(waveform)
|
157 |
|
158 |
-
|
159 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
160 |
|
|
|
|
|
|
|
161 |
with torch.no_grad():
|
162 |
waveforms_torch = torch.vstack(waveforms)
|
163 |
output = model(waveforms_torch)
|
@@ -169,8 +241,31 @@ def predict_on_section(client_name, timestamp, eq_lat, eq_lon, radius_km, source
|
|
169 |
p_max_confidence = np.min([p_phases[i::len(waveforms)].std() for i in range(len(waveforms))])
|
170 |
s_max_confidence = np.min([s_phases[i::len(waveforms)].std() for i in range(len(waveforms))])
|
171 |
|
172 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
173 |
for i in range(len(waveforms)):
|
|
|
174 |
current_P = p_phases[i::len(waveforms)]
|
175 |
current_S = s_phases[i::len(waveforms)]
|
176 |
|
@@ -188,17 +283,46 @@ def predict_on_section(client_name, timestamp, eq_lat, eq_lon, radius_km, source
|
|
188 |
ax[0].set_ylabel('Z')
|
189 |
|
190 |
ax[0].xaxis.set_major_formatter(mdates.DateFormatter('%H:%M:%S'))
|
191 |
-
ax[0].xaxis.set_major_locator(mdates.SecondLocator(interval=
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
192 |
|
|
|
|
|
|
|
|
|
|
|
193 |
ax[0].scatter(None, None, color='r', marker='|', label='P')
|
194 |
ax[0].scatter(None, None, color='b', marker='|', label='S')
|
195 |
ax[0].legend()
|
196 |
|
197 |
-
|
198 |
-
|
199 |
-
|
200 |
-
|
201 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
202 |
fig.canvas.draw();
|
203 |
image = np.array(fig.canvas.renderer.buffer_rgba())
|
204 |
plt.close(fig)
|
@@ -212,30 +336,54 @@ model = Onset_picker.load_from_checkpoint("./weights.ckpt",
|
|
212 |
model.eval()
|
213 |
|
214 |
with gr.Blocks() as demo:
|
215 |
-
gr.
|
216 |
-
|
217 |
-
|
218 |
-
|
219 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
220 |
""")
|
221 |
-
with gr.Tab("
|
222 |
-
|
223 |
-
|
224 |
-
|
225 |
-
|
226 |
-
|
227 |
-
|
228 |
-
|
229 |
-
|
230 |
-
|
|
|
|
|
|
|
231 |
|
232 |
button = gr.Button("Predict phases")
|
233 |
-
outputs = gr.
|
234 |
|
235 |
-
button.click(mark_phases, inputs=inputs, outputs=outputs)
|
236 |
|
237 |
with gr.Tab("Select earthquake from catalogue"):
|
238 |
-
gr.Markdown(
|
|
|
239 |
|
240 |
client_inputs = gr.Dropdown(
|
241 |
choices = list(URL_MAPPINGS.keys()),
|
@@ -244,6 +392,7 @@ with gr.Blocks() as demo:
|
|
244 |
value = "IRIS",
|
245 |
interactive=True
|
246 |
)
|
|
|
247 |
with gr.Row():
|
248 |
|
249 |
timestamp_inputs = gr.Textbox(value='2019-07-04 17:33:49',
|
@@ -277,13 +426,23 @@ with gr.Blocks() as demo:
|
|
277 |
interactive=True)
|
278 |
|
279 |
velocity_inputs = gr.Dropdown(
|
280 |
-
choices = ['1066a', '1066b', 'ak135',
|
|
|
|
|
281 |
label="1D velocity model",
|
282 |
info="Velocity model for station selection",
|
283 |
value = "1066a",
|
284 |
interactive=True
|
285 |
)
|
286 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
287 |
|
288 |
button = gr.Button("Predict phases")
|
289 |
outputs_section = gr.Image(label='Waveforms with Phases Marked', type='numpy', interactive=False)
|
@@ -291,13 +450,8 @@ with gr.Blocks() as demo:
|
|
291 |
button.click(predict_on_section,
|
292 |
inputs=[client_inputs, timestamp_inputs,
|
293 |
eq_lat_inputs, eq_lon_inputs,
|
294 |
-
radius_inputs, source_depth_inputs,
|
|
|
295 |
outputs=outputs_section)
|
296 |
|
297 |
-
with gr.Tab("Predict on your own waveform"):
|
298 |
-
gr.Markdown("""
|
299 |
-
Please upload your waveform in .npy (numpy) format.
|
300 |
-
Your waveform should be sampled at 100 sps and have 3 (Z, N, E) or 1 (Z) channels.
|
301 |
-
""")
|
302 |
-
|
303 |
demo.launch()
|
|
|
8 |
import torch
|
9 |
|
10 |
from scipy.stats import gaussian_kde
|
11 |
+
from bmi_topography import Topography
|
12 |
+
import earthpy.spatial as es
|
13 |
|
14 |
import obspy
|
15 |
from obspy.clients.fdsn import Client
|
|
|
22 |
|
23 |
import matplotlib.pyplot as plt
|
24 |
import matplotlib.dates as mdates
|
25 |
+
from matplotlib.colors import LightSource
|
26 |
|
27 |
from glob import glob
|
28 |
|
|
|
39 |
|
40 |
return processed_input, p_phase, s_phase
|
41 |
|
42 |
+
def mark_phases(waveform, uploaded_file):
|
43 |
+
|
44 |
+
if uploaded_file is not None:
|
45 |
+
waveform = uploaded_file.name
|
46 |
+
|
47 |
processed_input, p_phase, s_phase = make_prediction(waveform)
|
48 |
|
49 |
# Create a plot of the waveform with the phases marked
|
50 |
if sum(processed_input[0][2] == 0): #if input is 1C
|
51 |
fig, ax = plt.subplots(nrows=2, figsize=(10, 2), sharex=True)
|
52 |
|
53 |
+
ax[0].plot(processed_input[0][0], color='black', lw=1)
|
54 |
ax[0].set_ylabel('Norm. Ampl.')
|
55 |
|
56 |
else: #if input is 3C
|
57 |
fig, ax = plt.subplots(nrows=4, figsize=(10, 6), sharex=True)
|
58 |
+
ax[0].plot(processed_input[0][0], color='black', lw=1)
|
59 |
+
ax[1].plot(processed_input[0][1], color='black', lw=1)
|
60 |
+
ax[2].plot(processed_input[0][2], color='black', lw=1)
|
61 |
|
62 |
ax[0].set_ylabel('Z')
|
63 |
ax[1].set_ylabel('N')
|
|
|
89 |
plt.close(fig)
|
90 |
return image
|
91 |
|
92 |
+
def bin_distances(distances, bin_size=10):
|
93 |
+
# Bin the distances into groups of `bin_size` kilometers
|
94 |
+
binned_distances = {}
|
95 |
+
for i, distance in enumerate(distances):
|
96 |
+
bin_index = distance // bin_size
|
97 |
+
if bin_index not in binned_distances:
|
98 |
+
binned_distances[bin_index] = (distance, i)
|
99 |
+
elif i < binned_distances[bin_index][1]:
|
100 |
+
binned_distances[bin_index] = (distance, i)
|
101 |
+
|
102 |
+
# Select the first distance in each bin and its index
|
103 |
+
first_distances = []
|
104 |
+
for bin_index in binned_distances:
|
105 |
+
first_distance, first_distance_index = binned_distances[bin_index]
|
106 |
+
first_distances.append(first_distance_index)
|
107 |
+
|
108 |
+
return first_distances
|
109 |
+
|
110 |
def variance_coefficient(residuals):
|
111 |
# calculate the variance of the residuals
|
112 |
var = residuals.var()
|
|
|
113 |
# scale the variance to a coefficient between 0 and 1
|
114 |
coeff = 1 - (var / (residuals.max() - residuals.min()))
|
|
|
115 |
return coeff
|
116 |
|
117 |
+
def predict_on_section(client_name, timestamp, eq_lat, eq_lon, radius_km, source_depth_km, velocity_model, max_waveforms):
|
118 |
distances, t0s, st_lats, st_lons, waveforms = [], [], [], [], []
|
119 |
|
120 |
taup_model = TauPyModel(model=velocity_model)
|
121 |
client = Client(client_name)
|
122 |
|
123 |
window = radius_km / 111.2
|
124 |
+
max_waveforms = int(max_waveforms)
|
125 |
|
126 |
assert eq_lat - window > -90 and eq_lat + window < 90, "Latitude out of bounds"
|
127 |
assert eq_lon - window > -180 and eq_lon + window < 180, "Longitude out of bounds"
|
|
|
129 |
starttime = obspy.UTCDateTime(timestamp)
|
130 |
endtime = starttime + 120
|
131 |
|
132 |
+
try:
|
133 |
+
print('Starting to download inventory')
|
134 |
+
inv = client.get_stations(network="*", station="*", location="*", channel="*H*",
|
135 |
+
starttime=starttime, endtime=endtime,
|
136 |
+
minlatitude=(eq_lat-window), maxlatitude=(eq_lat+window),
|
137 |
+
minlongitude=(eq_lon-window), maxlongitude=(eq_lon+window),
|
138 |
+
level='station')
|
139 |
+
print('Finished downloading inventory')
|
140 |
+
except (IndexError, FDSNNoDataException, FDSNTimeoutException, FDSNInternalServerException):
|
141 |
+
fig, ax = plt.subplots()
|
142 |
+
ax.text(0.5,0.5,'Something is wrong with the data provider, try another')
|
143 |
+
fig.canvas.draw();
|
144 |
+
image = np.array(fig.canvas.renderer.buffer_rgba())
|
145 |
+
plt.close(fig)
|
146 |
+
return image
|
147 |
|
148 |
waveforms = []
|
149 |
cached_waveforms = glob("data/cached/*.mseed")
|
150 |
|
151 |
for network in inv:
|
152 |
+
# Skip the SYntetic networks
|
153 |
+
if network.code == 'SY':
|
154 |
+
continue
|
155 |
for station in network:
|
156 |
+
print(f"Processing {network.code}.{station.code}...")
|
157 |
+
distance = locations2degrees(eq_lat, eq_lon, station.latitude, station.longitude)
|
158 |
|
159 |
+
arrivals = taup_model.get_travel_times(source_depth_in_km=source_depth_km,
|
160 |
+
distance_in_degree=distance,
|
161 |
+
phase_list=["P", "S"])
|
162 |
|
163 |
+
if len(arrivals) > 0:
|
|
|
|
|
|
|
164 |
|
165 |
+
starttime = obspy.UTCDateTime(timestamp) + arrivals[0].time - 15
|
166 |
+
endtime = starttime + 60
|
167 |
+
try:
|
168 |
if f"data/cached/{network.code}_{station.code}_{starttime}.mseed" not in cached_waveforms:
|
169 |
+
print('Downloading waveform')
|
170 |
waveform = client.get_waveforms(network=network.code, station=station.code, location="*", channel="*",
|
171 |
starttime=starttime, endtime=endtime)
|
172 |
waveform.write(f"data/cached/{network.code}_{station.code}_{starttime}.mseed", format="MSEED")
|
173 |
+
print('Finished downloading and caching waveform')
|
174 |
else:
|
175 |
+
print('Reading cached waveform')
|
176 |
waveform = obspy.read(f"data/cached/{network.code}_{station.code}_{starttime}.mseed")
|
177 |
+
|
178 |
+
|
179 |
+
except (IndexError, FDSNNoDataException, FDSNTimeoutException, FDSNInternalServerException):
|
180 |
+
print(f'Skipping {network.code}_{station.code}_{starttime}')
|
181 |
+
continue
|
182 |
+
|
183 |
+
waveform = waveform.select(channel="H[BH][ZNE]")
|
184 |
+
waveform = waveform.merge(fill_value=0)
|
185 |
+
waveform = waveform[:3]
|
186 |
+
|
187 |
+
len_check = [len(x.data) for x in waveform]
|
188 |
+
if len(set(len_check)) > 1:
|
189 |
+
continue
|
190 |
+
|
191 |
+
if len(waveform) == 3:
|
192 |
+
try:
|
193 |
+
waveform = prepare_waveform(np.stack([x.data for x in waveform]))
|
194 |
|
|
|
|
|
|
|
|
|
|
|
|
|
195 |
distances.append(distance)
|
196 |
t0s.append(starttime)
|
197 |
st_lats.append(station.latitude)
|
198 |
st_lons.append(station.longitude)
|
199 |
waveforms.append(waveform)
|
200 |
|
201 |
+
print(f"Added {network.code}.{station.code} to the list of waveforms")
|
202 |
+
|
203 |
+
except:
|
204 |
+
continue
|
205 |
+
|
206 |
+
|
207 |
+
# If there are no waveforms, return an empty plot
|
208 |
+
if len(waveforms) == 0:
|
209 |
+
fig, ax = plt.subplots()
|
210 |
+
ax.text(0.5,0.5,'No waveforms found')
|
211 |
+
fig.canvas.draw();
|
212 |
+
image = np.array(fig.canvas.renderer.buffer_rgba())
|
213 |
+
plt.close(fig)
|
214 |
+
return image
|
215 |
+
|
216 |
+
|
217 |
+
first_distances = bin_distances(distances, bin_size=10/111.2)
|
218 |
+
|
219 |
+
# Edge case when there are way too many waveforms to process
|
220 |
+
selection_indexes = np.random.choice(first_distances,
|
221 |
+
np.min([len(first_distances), max_waveforms]),
|
222 |
+
replace=False)
|
223 |
+
|
224 |
+
waveforms = np.array(waveforms)[selection_indexes]
|
225 |
+
distances = np.array(distances)[selection_indexes]
|
226 |
+
t0s = np.array(t0s)[selection_indexes]
|
227 |
+
st_lats = np.array(st_lats)[selection_indexes]
|
228 |
+
st_lons = np.array(st_lons)[selection_indexes]
|
229 |
|
230 |
+
waveforms = [torch.tensor(waveform) for waveform in waveforms]
|
231 |
+
|
232 |
+
print('Starting to run predictions')
|
233 |
with torch.no_grad():
|
234 |
waveforms_torch = torch.vstack(waveforms)
|
235 |
output = model(waveforms_torch)
|
|
|
241 |
p_max_confidence = np.min([p_phases[i::len(waveforms)].std() for i in range(len(waveforms))])
|
242 |
s_max_confidence = np.min([s_phases[i::len(waveforms)].std() for i in range(len(waveforms))])
|
243 |
|
244 |
+
print(f"Starting plotting {len(waveforms)} waveforms")
|
245 |
+
fig, ax = plt.subplots(nrows=1, ncols=3, figsize=(10, 3))
|
246 |
+
|
247 |
+
# Plot topography
|
248 |
+
print('Fetching topography')
|
249 |
+
params = Topography.DEFAULT.copy()
|
250 |
+
extra_window = 0.5
|
251 |
+
params["south"] = np.min([st_lats.min(), eq_lat])-extra_window
|
252 |
+
params["north"] = np.max([st_lats.max(), eq_lat])+extra_window
|
253 |
+
params["west"] = np.min([st_lons.min(), eq_lon])-extra_window
|
254 |
+
params["east"] = np.max([st_lons.max(), eq_lon])+extra_window
|
255 |
+
|
256 |
+
topo_map = Topography(**params)
|
257 |
+
topo_map.fetch()
|
258 |
+
topo_map.load()
|
259 |
+
|
260 |
+
print('Plotting topo')
|
261 |
+
hillshade = es.hillshade(topo_map.da[0], altitude=10)
|
262 |
+
|
263 |
+
topo_map.da.plot(ax = ax[1], cmap='Greys', add_colorbar=False, add_labels=False)
|
264 |
+
topo_map.da.plot(ax = ax[2], cmap='Greys', add_colorbar=False, add_labels=False)
|
265 |
+
ax[1].imshow(hillshade, cmap="Greys", alpha=0.5)
|
266 |
+
|
267 |
for i in range(len(waveforms)):
|
268 |
+
print(f"Plotting waveform {i+1}/{len(waveforms)}")
|
269 |
current_P = p_phases[i::len(waveforms)]
|
270 |
current_S = s_phases[i::len(waveforms)]
|
271 |
|
|
|
283 |
ax[0].set_ylabel('Z')
|
284 |
|
285 |
ax[0].xaxis.set_major_formatter(mdates.DateFormatter('%H:%M:%S'))
|
286 |
+
ax[0].xaxis.set_major_locator(mdates.SecondLocator(interval=20))
|
287 |
+
|
288 |
+
delta_t = t0s[i].timestamp - obspy.UTCDateTime(timestamp).timestamp
|
289 |
+
|
290 |
+
velocity_p = (distances[i]*111.2)/(delta_t+current_P.mean()*60).item()
|
291 |
+
velocity_s = (distances[i]*111.2)/(delta_t+current_S.mean()*60).item()
|
292 |
+
|
293 |
+
print(f"Station {st_lats[i]}, {st_lons[i]} has P velocity {velocity_p} and S velocity {velocity_s}")
|
294 |
+
|
295 |
+
# Generate an array from st_lat to eq_lat and from st_lon to eq_lon
|
296 |
+
x = np.linspace(st_lons[i], eq_lon, 50)
|
297 |
+
y = np.linspace(st_lats[i], eq_lat, 50)
|
298 |
|
299 |
+
# Plot the array
|
300 |
+
ax[1].scatter(x, y, c=np.zeros_like(x)+velocity_p, alpha=0.5, vmin=0, vmax=8)
|
301 |
+
ax[2].scatter(x, y, c=np.zeros_like(x)+velocity_s, alpha=0.5, vmin=0, vmax=8)
|
302 |
+
|
303 |
+
# Add legend
|
304 |
ax[0].scatter(None, None, color='r', marker='|', label='P')
|
305 |
ax[0].scatter(None, None, color='b', marker='|', label='S')
|
306 |
ax[0].legend()
|
307 |
|
308 |
+
print('Plotting stations')
|
309 |
+
for i in range(1,3):
|
310 |
+
ax[i].scatter(st_lons, st_lats, color='b', label='Stations')
|
311 |
+
ax[i].scatter(eq_lon, eq_lat, color='r', marker='*', label='Earthquake')
|
312 |
+
|
313 |
+
# Generate colorbar for the velocity plot
|
314 |
+
cbar = plt.colorbar(ax[1].scatter(None, None, c=velocity_p, alpha=0.5, vmin=0, vmax=8), ax=ax[1])
|
315 |
+
cbar.set_label('P Velocity (km/s)')
|
316 |
+
ax[1].set_title('P Velocity')
|
317 |
+
|
318 |
+
cbar = plt.colorbar(ax[2].scatter(None, None, c=velocity_s, alpha=0.5, vmin=0, vmax=8), ax=ax[2])
|
319 |
+
cbar.set_label('S Velocity (km/s)')
|
320 |
+
ax[2].set_title('S Velocity')
|
321 |
+
|
322 |
+
|
323 |
+
|
324 |
+
plt.subplots_adjust(hspace=0., wspace=0.5)
|
325 |
+
|
326 |
fig.canvas.draw();
|
327 |
image = np.array(fig.canvas.renderer.buffer_rgba())
|
328 |
plt.close(fig)
|
|
|
336 |
model.eval()
|
337 |
|
338 |
with gr.Blocks() as demo:
|
339 |
+
gr.HTML("""<h1>PhaseHunter</h1>
|
340 |
+
<p>This app allows one to detect <mark style="background-color: red; color: white;">P</mark> and <mark style="background-color: blue; color: white;">S</mark> seismic phases along with
|
341 |
+
|
342 |
+
<span style="font-size: 24px; font-weight: bold;">
|
343 |
+
<span style="color: #4c4c4c; text-shadow: 1px 1px 0 #ccc, -1px -1px 0 #ccc, 1px -1px 0 #ccc, -1px 1px 0 #ccc;">u</span>
|
344 |
+
<span style="color: #808080; text-shadow: 1px 1px 0 #bbb, -1px -1px 0 #bbb, 1px -1px 0 #bbb, -1px 1px 0 #bbb;">n</span>
|
345 |
+
<span style="color: #b3b3b3; text-shadow: 1px 1px 0 #aaa, -1px -1px 0 #aaa, 1px -1px 0 #aaa, -1px 1px 0 #aaa;">c</span>
|
346 |
+
<span style="color: #d9d9d9; text-shadow: 1px 1px 0 #ccc, -1px -1px 0 #ccc, 1px -1px 0 #ccc, -1px 1px 0 #ccc;">e</span>
|
347 |
+
<span style="color: #f2f2f2; text-shadow: 1px 1px 0 #eee, -1px -1px 0 #eee, 1px -1px 0 #eee, -1px 1px 0 #eee;">r</span>
|
348 |
+
<span style="color: #d9d9d9; text-shadow: 1px 1px 0 #ccc, -1px -1px 0 #ccc, 1px -1px 0 #ccc, -1px 1px 0 #ccc;">t</span>
|
349 |
+
<span style="color: #b3b3b3; text-shadow: 1px 1px 0 #aaa, -1px -1px 0 #aaa, 1px -1px 0 #aaa, -1px 1px 0 #aaa;">a</span>
|
350 |
+
<span style="color: #808080; text-shadow: 1px 1px 0 #bbb, -1px -1px 0 #bbb, 1px -1px 0 #bbb, -1px 1px 0 #bbb;">i</span>
|
351 |
+
<span style="color: #4c4c4c; text-shadow: 1px 1px 0 #ccc, -1px -1px 0 #ccc, 1px -1px 0 #ccc, -1px 1px 0 #ccc;">n</span>
|
352 |
+
<span style="color: #808080; text-shadow: 1px 1px 0 #bbb, -1px -1px 0 #bbb, 1px -1px 0 #bbb, -1px 1px 0 #bbb;">t</span>
|
353 |
+
<span style="color: #b3b3b3; text-shadow: 1px 1px 0 #aaa, -1px -1px 0 #aaa, 1px -1px 0 #aaa, -1px 1px 0 #aaa;">y</span>
|
354 |
+
|
355 |
+
</span>
|
356 |
+
of the detection.</p>
|
357 |
+
<ol>
|
358 |
+
<li>By selecting one of the sample waveforms.</li>
|
359 |
+
<li>By uploading your own waveform.</li>
|
360 |
+
<li>By selecting an earthquake from the global earthquake catalogue.</li>
|
361 |
+
</ol>
|
362 |
+
<p>Please upload your waveform in <code>.npy</code> (numpy) format.</p>
|
363 |
+
<p>Your waveform should be sampled at 100 samples per second and have 3 (Z, N, E) or 1 (Z) channels. If your file is longer than 60 seconds, the app will only use the first 60 seconds of the waveform.</p>
|
364 |
""")
|
365 |
+
with gr.Tab("Try on a single station"):
|
366 |
+
with gr.Row():
|
367 |
+
# Define the input and output types for Gradio
|
368 |
+
inputs = gr.Dropdown(
|
369 |
+
["data/sample/sample_0.npy",
|
370 |
+
"data/sample/sample_1.npy",
|
371 |
+
"data/sample/sample_2.npy"],
|
372 |
+
label="Sample waveform",
|
373 |
+
info="Select one of the samples",
|
374 |
+
value = "data/sample/sample_0.npy"
|
375 |
+
)
|
376 |
+
|
377 |
+
upload = gr.File(label="Or upload your own waveform")
|
378 |
|
379 |
button = gr.Button("Predict phases")
|
380 |
+
outputs = gr.Image(label='Waveform with Phases Marked', type='numpy', interactive=False)
|
381 |
|
382 |
+
button.click(mark_phases, inputs=[inputs, upload], outputs=outputs)
|
383 |
|
384 |
with gr.Tab("Select earthquake from catalogue"):
|
385 |
+
gr.Markdown("""Select an earthquake from the global earthquake catalogue and the app will download the waveform from the FDSN client of your choice.
|
386 |
+
""")
|
387 |
|
388 |
client_inputs = gr.Dropdown(
|
389 |
choices = list(URL_MAPPINGS.keys()),
|
|
|
392 |
value = "IRIS",
|
393 |
interactive=True
|
394 |
)
|
395 |
+
|
396 |
with gr.Row():
|
397 |
|
398 |
timestamp_inputs = gr.Textbox(value='2019-07-04 17:33:49',
|
|
|
426 |
interactive=True)
|
427 |
|
428 |
velocity_inputs = gr.Dropdown(
|
429 |
+
choices = ['1066a', '1066b', 'ak135',
|
430 |
+
'ak135f', 'herrin', 'iasp91',
|
431 |
+
'jb', 'prem', 'pwdk'],
|
432 |
label="1D velocity model",
|
433 |
info="Velocity model for station selection",
|
434 |
value = "1066a",
|
435 |
interactive=True
|
436 |
)
|
437 |
+
|
438 |
+
max_waveforms_inputs = gr.Slider(minimum=1,
|
439 |
+
maximum=100,
|
440 |
+
value=10,
|
441 |
+
label="Max waveforms per section",
|
442 |
+
step=1,
|
443 |
+
info="Maximum number of waveforms to show per section\n (to avoid long prediction times)",
|
444 |
+
interactive=True,
|
445 |
+
)
|
446 |
|
447 |
button = gr.Button("Predict phases")
|
448 |
outputs_section = gr.Image(label='Waveforms with Phases Marked', type='numpy', interactive=False)
|
|
|
450 |
button.click(predict_on_section,
|
451 |
inputs=[client_inputs, timestamp_inputs,
|
452 |
eq_lat_inputs, eq_lon_inputs,
|
453 |
+
radius_inputs, source_depth_inputs,
|
454 |
+
velocity_inputs, max_waveforms_inputs],
|
455 |
outputs=outputs_section)
|
456 |
|
|
|
|
|
|
|
|
|
|
|
|
|
457 |
demo.launch()
|
data/.DS_Store
CHANGED
Binary files a/data/.DS_Store and b/data/.DS_Store differ
|
|
data/cached/BC_JARAX_2023-04-01T01:16:13.997267Z.mseed
ADDED
Binary file (73.7 kB). View file
|
|
data/cached/CE_24944_2023-04-01T01:16:14.044529Z.mseed
ADDED
Binary file (28.7 kB). View file
|
|
data/cached/CE_24945_2023-04-01T01:16:13.506381Z.mseed
ADDED
Binary file (28.7 kB). View file
|
|
data/cached/CI_ADO_2019-07-04T17:33:53.650962Z.mseed
ADDED
Binary file (154 kB). View file
|
|
data/cached/CI_ARV_2019-07-04T17:33:53.096986Z.mseed
ADDED
Binary file (124 kB). View file
|
|
data/cached/CI_CJV2_2023-04-01T01:16:16.011425Z.mseed
ADDED
Binary file (188 kB). View file
|
|
data/cached/CI_CWC_2019-07-04T17:33:47.189005Z.mseed
ADDED
Binary file (154 kB). View file
|
|
data/cached/CI_EDW2_2019-07-04T17:33:49.567241Z.mseed
ADDED
Binary file (143 kB). View file
|
|
data/cached/CI_FUR_2019-07-04T17:33:49.310305Z.mseed
ADDED
Binary file (147 kB). View file
|
|
data/cached/CI_GRA_2019-07-04T17:33:53.959922Z.mseed
ADDED
Binary file (133 kB). View file
|
|
data/cached/CI_GSC_2019-07-04T17:33:47.536363Z.mseed
ADDED
Binary file (155 kB). View file
|
|
data/cached/CI_HEC_2019-07-04T17:33:56.148977Z.mseed
ADDED
Binary file (140 kB). View file
|
|
data/cached/CI_ISA_2019-07-04T17:33:46.297658Z.mseed
ADDED
Binary file (148 kB). View file
|
|
data/cached/CI_JNH2_2023-04-01T01:16:13.664044Z.mseed
ADDED
Binary file (180 kB). View file
|
|
data/cached/CI_JRC2_2019-07-04T17:33:39.947494Z.mseed
ADDED
Binary file (225 kB). View file
|
|
data/cached/CI_LRL_2019-07-04T17:33:40.248999Z.mseed
ADDED
Binary file (164 kB). View file
|
|
data/cached/CI_LRR2_2023-04-01T01:16:15.095101Z.mseed
ADDED
Binary file (180 kB). View file
|
|
data/cached/CI_MPM_2019-07-04T17:33:40.443346Z.mseed
ADDED
Binary file (124 kB). View file
|
|
data/cached/CI_OSI_2019-07-04T17:33:57.203547Z.mseed
ADDED
Binary file (131 kB). View file
|
|
data/cached/CI_PUT_2023-04-01T01:16:16.558724Z.mseed
ADDED
Binary file (332 kB). View file
|
|
data/cached/CI_Q0013_2019-07-04T17:33:54.489098Z.mseed
ADDED
Binary file (75.3 kB). View file
|
|
data/cached/CI_Q0024_2019-07-04T17:33:55.620507Z.mseed
ADDED
Binary file (75.8 kB). View file
|
|
data/cached/CI_Q0035_2019-07-04T17:33:56.041452Z.mseed
ADDED
Binary file (75.3 kB). View file
|
|
data/cached/CI_Q0056_2019-07-04T17:33:50.407027Z.mseed
ADDED
Binary file (86.5 kB). View file
|
|
data/cached/CI_Q0061_2019-07-04T17:33:53.114695Z.mseed
ADDED
Binary file (81.9 kB). View file
|
|
data/cached/CI_Q0068_2019-07-04T17:33:46.411249Z.mseed
ADDED
Binary file (80.9 kB). View file
|
|
data/cached/CI_Q0072_2019-07-04T17:33:38.390421Z.mseed
ADDED
Binary file (83.5 kB). View file
|
|
data/cached/CI_RRX_2019-07-04T17:33:50.712219Z.mseed
ADDED
Binary file (144 kB). View file
|
|
data/cached/CI_SBB2_2023-04-01T01:16:15.609344Z.mseed
ADDED
Binary file (180 kB). View file
|
|
data/cached/CI_SHO_2019-07-04T17:33:51.673022Z.mseed
ADDED
Binary file (147 kB). View file
|
|
data/cached/CI_SLA_2019-07-04T17:33:40.190746Z.mseed
ADDED
Binary file (151 kB). View file
|
|
data/cached/CI_SRA_2023-04-01T01:16:14.887472Z.mseed
ADDED
Binary file (180 kB). View file
|
|
data/cached/CI_SRT_2019-07-04T17:33:38.029990Z.mseed
ADDED
Binary file (221 kB). View file
|
|
data/cached/CI_TIN_2019-07-04T17:33:55.946998Z.mseed
ADDED
Binary file (135 kB). View file
|
|
data/cached/CI_TOW2_2019-07-04T17:33:37.991033Z.mseed
ADDED
Binary file (229 kB). View file
|
|
data/cached/CI_VCS_2023-04-01T01:16:15.307897Z.mseed
ADDED
Binary file (180 kB). View file
|
|
data/cached/CI_VTV_2019-07-04T17:33:53.688913Z.mseed
ADDED
Binary file (153 kB). View file
|
|
data/cached/CI_WBM_2019-07-04T17:33:40.063644Z.mseed
ADDED
Binary file (225 kB). View file
|
|
data/cached/CI_WLS2_2023-04-01T01:16:13.623472Z.mseed
ADDED
Binary file (319 kB). View file
|
|
data/cached/CI_WMF_2019-07-04T17:33:41.867962Z.mseed
ADDED
Binary file (217 kB). View file
|
|
data/cached/CI_WRC2_2019-07-04T17:33:38.698107Z.mseed
ADDED
Binary file (229 kB). View file
|
|
data/cached/CI_WRV2_2019-07-04T17:33:40.843660Z.mseed
ADDED
Binary file (94.2 kB). View file
|
|
data/cached/CI_WVP2_2019-07-04T17:33:39.650418Z.mseed
ADDED
Binary file (94.2 kB). View file
|
|
data/cached/LB_DAC_2019-07-04T17:33:43.387184Z.mseed
ADDED
Binary file (115 kB). View file
|
|
data/cached/NN_FMT_2019-07-04T17:33:51.798341Z.mseed
ADDED
Binary file (30.2 kB). View file
|
|
data/cached/NN_GVN_2019-07-04T17:33:54.053591Z.mseed
ADDED
Binary file (30.2 kB). View file
|
|