def _choose_connection(**datafind_kw):
if os.getenv('LIGO_DATAFIND_SERVER') or datafind_kw.get('host'):
from gwdatafind import connect
return connect(**datafind_kw)
if os.getenv('VIRGODATA'):
return FflConnection()
raise RuntimeError("unknown datafind configuration, cannot discover data")
def datafind_connection(server=None):
""" Return a connection to the datafind server
Parameters
-----------
server : {SERVER:PORT, string}, optional
A string representation of the server and port.
The port may be ommitted.
Returns
--------
connection
The open connection to the datafind server.
"""
return gwdatafind.connect(host=server)
def initGWDataFind(enforce=False):
global GWDataFindInitialized
if not GWDataFindInitialized or enforce:
fastlog(INFO,"Initializing GWDataFind connection.")
if shouldRenew():
fastlog(INFO, "Creating new VOMS proxy or VO=virgo:virgo/virgo and 48h validity.")
generateVomsProxy("virgo:virgo/virgo", 48)
global GWDataFindConn
GWDataFindConn = connect(GWDataFindServer)
global GWDataFindObservatories
GWDataFindObservatories = GWDataFindConn.find_observatories()
global GWDataFindTypes
GWDataFindTypes = {}
for obs in GWDataFindObservatories:
GWDataFindTypes[obs] = GWDataFindConn.find_types(obs)
GWDataFindInitialized = True
# Reads the save channels for L1_o3b | file found here: https://dcc.ligo.org/T2000277
safe_channels = pd.read_csv("L1_o3b_safe.csv", usecols=['Channel']).Channel.to_list()
safe_ISI_list = [chan for chan in safe_channels if 'L1:ISI' in chan] # only ISI channels (seismic)
for start, end, filetag in data_times:
# create output file for the segments | will override existing files
with h5py.File(f'./segments/aux_{filetag}.hdf5', 'w') as f:
meta = f.create_group('meta')
meta.create_dataset('Start GPS', data=int(start))
# End GPS will be saved at the end of the script
# uses gwdatafind to find the locations of the frame files (CVMFS) | https://gwdatafind.readthedocs.io/en/stable/index.html
# extra info : https://gwpy.github.io/docs/v0.1/timeseries/gwf.html
# https://computing.docs.ligo.org/guide/data/#cvmfs-data-discovery
conn = gwdatafind.connect()
urls = conn.find_urls("L", "L1_R", start, end) # L1_R : All 'raw' data channels, stored at the native sampling rate
conn.close() # not sure if needed
prev_file_data = None # will hold left over data from previous file that did not fit into a segment
for url in tqdm(urls):
frame_start = int(url.split('-')[4])
frame_size = int(url.split('-')[5].split('.')[0])
# read data of all channeles in 'safe_ISI_list' from frame file located at 'url'
data_dict = gwpy.timeseries.TimeSeriesDict.read(url, safe_ISI_list , start=frame_start, end=frame_start+frame_size) # start, end possibly not needed?
# 2d lists to store data and channel names (tags) sorted by their sampling rates | sort is necessary for calculation of fractal dimension
data_all = [[], [], [], [], []]
tags = [[], [], [], [], []]
for key in data_dict.keys(): # loop over all aux channels and fill 2d lists
idx = int(math.log(data_dict[key].sample_rate.value, 2))-8 # convert sampling rate to index for 2d lists
def _choose_connection(**datafind_kw):
if os.getenv('LIGO_DATAFIND_SERVER') or datafind_kw.get('host'):
return connect(**datafind_kw)
if os.getenv('VIRGODATA'):
return FflConnection()
raise RuntimeError("unknown datafind configuration, cannot discover data")