Skip to content

Latest commit

 

History

History
271 lines (215 loc) · 8.28 KB

README.md

File metadata and controls

271 lines (215 loc) · 8.28 KB

TXED

Description

TXED is the Texas Earthquake Dataset for AI

Reference

Chen, Y., A. Savvaidis, O. M. Saad, G.-C. D. Huang, D. Siervo, V. O’Sullivan, C. McCabe, B. Uku, P. Fleck, G. Burke, N. L. Alvarez, J. Domino, and I. Grigoratos, “TXED: the texas earthquake dataset for AI,” Seismological Research Letters, vol. 1, no. 1, p. doi: 10.1785/0220230327, 2024.

BibTeX:

@Article{txed,
  author={Yangkang Chen and Alexandros Savvaidis and Omar M. Saad and Guo-Chin Dino Huang and Daniel Siervo and Vincent O’Sullivan and Cooper McCabe and Bede Uku and Preston Fleck and Grace Burke and Natalie L. Alvarez and Jessica Domino and Iason Grigoratos},
  title = {{TXED}: the Texas Earthquake Dataset for {AI}},
  journal={Seismological Research Letters},
  year=2024,
  volume=1,
  issue=1,
  number=1,
  pages={doi: 10.1785/0220230327},
  doi={10.1785/0220230327},
}

Copyright

Developers of the TXED package, 2021-present

License

GNU General Public License, Version 3
(http://www.gnu.org/copyleft/gpl.html)

Install

Using the latest version

git clone https://github.com/chenyk1990/txed

Download

Google Drive link: https://drive.google.com/drive/folders/1WXVB8ytNB4bOaZ97oq6OmMRyAEg95trp?usp=sharing

wget /address_TBD/TXED_20231111.h5
wget /address_TBD/ID_20231111.npy

Examples

Check the INFO of signal waveforms

import numpy as np
allid = np.load("ID_20231111.npy")
signalid=[ii for ii in allid if ii.split("_")[-1]=='EV']
print('Length of signalid is',len(signalid))

Check the INFO of noise waveforms

import numpy as np
allid = np.load("ID_20231111.npy")
noiseid=[ii for ii in allid if ii.split("_")[-1]=='NO']
print('Length of noiseid is',len(noiseid))

Print attributes in TXED

import h5py
import numpy as np
f = h5py.File("TXED_20231111.h5", 'r')
eventid=np.load("ID_20231111.npy")
idx=eventid[0]
dataset = f.get(idx)
print('TXED attributes are:',dataset.attrs.keys())

Plot signal waveforms

import h5py,os
import numpy as np
import matplotlib.pyplot as plt
plt.rcParams["figure.figsize"] = (7,8.2)

if os.path.isdir('./waveforms') == False:  
	os.makedirs('./waveforms',exist_ok=True)

h5fname="TXED_20231111.h5"
npyfname="ID_20231111.npy"

#open the h5file
f = h5py.File(h5fname, 'r')
eventid=np.load(npyfname)

#specificy a number of waveforms to plot
no=5
idxs=eventid[0:no]

#or specify an arbitrary TexNet eventID
# idxs=[ii for ii in eventid if ii.split('_')[0]=='texnet2023qnms']
# idxs=[ii for ii in eventid if ii.split('_')[0]=='texnet2022wmmd']
idxs=[ii for ii in eventid if ii.split('_')[0]=='texnet2023ncwh']

#loop over IDs and plot
id=0
for idx in idxs:
	id=id+1
	print('Plotting: %d/%d'%(id,len(idxs)))
	
	dataset = f.get(idx)
	data = np.array(dataset['data'])
	spt = int(dataset.attrs['p_arrival_sample']);
	sst = int(dataset.attrs['s_arrival_sample']);
	coda_end = int(dataset.attrs['coda_end_sample']);
	snr = dataset.attrs['snr_db'];
	t=dataset.attrs['origin_time']
	mag=dataset.attrs['magnitude']
	
	fig=plt.figure()
	ax1 = fig.add_subplot(311)
	plt.plot(data[:,0], 'k',label='Z')
	ymin,yma = ax1.get_ylim()
	plt.vlines(spt,ymin,yma,color='r',linewidth=2)
	plt.vlines(sst,ymin,yma,color='b',linewidth=2)
	legend_properties = {'weight':'bold'}
	ymin, ymax = ax1.get_ylim()
	plt.legend(loc = 'upper right', borderaxespad=0., prop=legend_properties)
	plt.ylabel('Amplitude', fontsize=12) 
	ax1.set_xticklabels([])
	ax = fig.add_subplot(312) 
	plt.plot(data[:,1], 'k',label='N')
	ymin,yma = ax1.get_ylim()
	plt.vlines(spt,ymin,yma,color='r',linewidth=2)
	plt.vlines(sst,ymin,yma,color='b',linewidth=2)
	legend_properties = {'weight':'bold'}
	ymin, ymax = ax.get_ylim()
	plt.legend(loc = 'upper right', borderaxespad=0., prop=legend_properties)
	plt.ylabel('Amplitude', fontsize=12) 
	ax.set_xticklabels([])
	ax = fig.add_subplot(313) 
	plt.plot(data[:,2], 'k',label='E')
	ymin,yma = ax1.get_ylim()
	plt.vlines(spt,ymin,yma,color='r',linewidth=2)
	plt.vlines(sst,ymin,yma,color='b',linewidth=2)
	legend_properties = {'weight':'bold'}
	ymin, ymax = ax.get_ylim()
	plt.legend(loc = 'upper right', borderaxespad=0., prop=legend_properties)
	plt.ylabel('Amplitude', fontsize=12) 
	plt.xlabel('Sample', fontsize=12) 

	eid=idx.split("_")[0]
	stname=idx.split("_")[1]
	sttime=t
	plt.text(-1200,(ymin-(ymax-ymin)*0.3),eid+' (Ml=%.2g)'%mag,fontsize=12,color='k')
	plt.text(1800,(ymin-(ymax-ymin)*0.3),stname,fontsize=12,color='k')
	plt.text(3700,(ymin-(ymax-ymin)*0.3),sttime,fontsize=12,color='k')
	ax1.set_title('Signal waveform: %s-%s'%(eid,stname), fontsize=14)
	plt.savefig(fname='./waveforms/signal-%s-%s'%(eid,stname)+'.png', format="png")
	plt.show()
	plt.close() 

f.close()

Plot noise waveforms

import h5py,os,random
import numpy as np
import matplotlib.pyplot as plt
plt.rcParams["figure.figsize"] = (7,8.2)

if os.path.isdir('./waveforms') == False:  
	os.makedirs('./waveforms',exist_ok=True)

h5fname="TXED_20231111.h5"
npyfname="ID_20231111.npy"

#open the h5file
f = h5py.File(h5fname, 'r')
allid=np.load(npyfname)

noiseid=[ii for ii in allid if ii.split("_")[-1]=='NO']
print('Length of noiseid is',len(noiseid))

#specificy a number of waveforms to plot
no=100

#random shuffling the noise waveforms
random.seed(2011)
random.shuffle(noiseid)
idxs=noiseid[0:no]

id=0
for idx in idxs:
	id=id+1
	print('Plotting: %d/%d'%(id,len(idxs)))
	dataset = f.get(idx)
	data = np.array(dataset['data'])
	t=dataset.attrs['origin_time']
	fig=plt.figure()
	ax1 = fig.add_subplot(311)
	plt.plot(data[:,0], 'k',label='Z')
	ymin,yma = ax1.get_ylim()
	legend_properties = {'weight':'bold'}
	ymin, ymax = ax1.get_ylim()
	plt.legend(loc = 'upper right', borderaxespad=0., prop=legend_properties)
	plt.ylabel('Amplitude', fontsize=12) 
	ax1.set_xticklabels([])
	ax = fig.add_subplot(312) 
	plt.plot(data[:,1], 'k',label='N')
	ymin,yma = ax1.get_ylim()
	legend_properties = {'weight':'bold'}
	ymin, ymax = ax.get_ylim()
	plt.legend(loc = 'upper right', borderaxespad=0., prop=legend_properties)
	plt.ylabel('Amplitude', fontsize=12) 
	ax.set_xticklabels([])
	ax = fig.add_subplot(313) 
	plt.plot(data[:,2], 'k',label='E')
	ymin,yma = ax1.get_ylim()
	legend_properties = {'weight':'bold'}
	ymin, ymax = ax.get_ylim()
	plt.legend(loc = 'upper right', borderaxespad=0., prop=legend_properties)
	plt.ylabel('Amplitude', fontsize=12) 
	plt.xlabel('Sample', fontsize=12) 

	no=idx.split("_")[0]
	eid='Noise #%s'%no
	stname=idx.split("_")[1]
	sttime=t
	plt.text(-1200,(ymin-(ymax-ymin)*0.3),eid,fontsize=12,color='k')
	plt.text(1200,(ymin-(ymax-ymin)*0.3),stname,fontsize=12,color='k')
	plt.text(3300,(ymin-(ymax-ymin)*0.3),sttime,fontsize=12,color='k')
	ax1.set_title('Noise waveform: #%s-%s'%(no,stname), fontsize=14)
	plt.savefig(fname='./waveforms/noise-%s-%s'%(no,stname)+'.png', format="png")
	plt.show()
	plt.close() 
f.close()

The Ipython Notebooks are examples for playing with the TXED.

Single-station location example

Development

The development team welcomes voluntary contributions from any open-source enthusiast. 
If you want to make contribution to this project, feel free to contact the development team.

Contact

Regarding any questions, bugs, developments, collaborations, please contact  
Yangkang Chen
[email protected]

NOTES:

Gallery

The gallery figures of the txed package can be found at https://github.com/chenyk1990/gallery/tree/main/txed

These gallery figures are also presented below.

A sample signal waveform Generated by test_signal.py Slicing

A sample noise waveform Generated by test_noise.py Slicing

Waveforms of an arbitrary TexNet event extracted from TXED Generated by test_event.py Slicing