pylot/pylot/core/util/dataprocessing.py

718 lines
27 KiB
Python

#!/usr/bin/env python
# -*- coding: utf-8 -*-
import glob
import logging
import os
import sys
import numpy as np
from obspy import UTCDateTime, read_inventory, read
from obspy.io.xseed import Parser
from pylot.core.util.utils import key_for_set_value, find_in_list, \
gen_Pool
class Metadata(object):
def __init__(self, inventory=None, verbosity=1):
self.inventories = []
# saves read metadata objects (Parser/inventory) for a filename
self.inventory_files = {}
# saves filenames holding metadata for a seed_id
# seed id as key, path to file as value
self.seed_ids = {}
self.stations_dict = {}
# saves which metadata files are from obspy dmt
self.obspy_dmt_invs = []
if inventory:
# make sure that no accidental backslashes mess up the path
if isinstance(inventory, str):
inventory = inventory.replace('\\', '/')
inventory = os.path.abspath(inventory)
if os.path.isdir(inventory):
self.add_inventory(inventory)
if os.path.isfile(inventory):
self.add_inventory_file(inventory)
self.verbosity = verbosity
def __str__(self):
repr = 'PyLoT Metadata object including the following inventories:\n\n'
ntotal = len(self.inventories)
for index, inventory in enumerate(self.inventories):
if index < 2 or (ntotal - index) < 3:
repr += '{}\n'.format(inventory)
if ntotal > 4 and int(ntotal / 2) == index:
repr += '...\n'
if ntotal > 4:
repr += '\nTotal of {} inventories. Use Metadata.inventories to see all.'.format(ntotal)
return repr
def __repr__(self):
return self.__str__()
def add_inventory(self, path_to_inventory, obspy_dmt_inv=False):
"""
Add path to list of inventories.
:param path_to_inventory: Path to a folder
:type path_to_inventory: str
:return: None
"""
path_to_inventory = path_to_inventory.replace('\\', '/')
path_to_inventory = os.path.abspath(path_to_inventory)
assert (os.path.isdir(path_to_inventory)), '{} is no directory'.format(path_to_inventory)
if path_to_inventory not in self.inventories:
self.inventories.append(path_to_inventory)
if obspy_dmt_inv == True:
self.obspy_dmt_invs.append(path_to_inventory)
def add_inventory_file(self, path_to_inventory_file):
"""
Add the folder in which the file exists to the list of inventories.
:param path_to_inventory_file: full path including filename
:type path_to_inventory_file: str
:return: None
"""
assert (os.path.isfile(path_to_inventory_file)), '{} is no file'.format(path_to_inventory_file)
self.add_inventory(os.path.split(path_to_inventory_file)[0])
if path_to_inventory_file not in self.inventory_files.keys():
self.read_single_file(path_to_inventory_file)
def remove_all_inventories(self):
self.__init__()
def remove_inventory(self, path_to_inventory):
"""
Remove a path from inventories list. If path is not in inventories list, do nothing.
:param path_to_inventory: Path to a folder
"""
if not path_to_inventory in self.inventories:
print('Path {} not in inventories list.'.format(path_to_inventory))
return
self.inventories.remove(path_to_inventory)
for filename in list(self.inventory_files.keys()):
if filename.startswith(path_to_inventory):
del (self.inventory_files[filename])
for seed_id in list(self.seed_ids.keys()):
if self.seed_ids[seed_id].startswith(path_to_inventory):
del (self.seed_ids[seed_id])
# have to clean self.stations_dict as well
# this will be rebuilt for the next init of the arraymap anyway, so just reset it
self.stations_dict = {}
def clear_inventory(self):
for inv in self.obspy_dmt_invs:
self.remove_inventory(inv)
self.obspy_dmt_invs = []
def get_metadata(self, seed_id, time=None):
"""
Get metadata for seed id at time. When time is not specified, metadata for current time is fetched.
:param seed_id: Seed id such as BW.WETR..HHZ (Network.Station.Location.Channel)
:type seed_id: str
:param time: Time for which the metadata should be returned
:type time: UTCDateTime
:return: Dictionary with keys data and invtype.
data is a obspy.io.xseed.parser.Parser or an obspy.core.inventory.inventory.Inventory depending on the metadata
file.
invtype is a string denoting of which type the value of the data key is. It can take the values 'dless',
'dseed', 'xml', 'resp', according to the filetype of the metadata.
:rtype: dict
"""
# try most recent data if no time is specified
if not time:
time = UTCDateTime()
# get metadata for a specific seed_id, if not already read, try to read from inventories
if not seed_id in self.seed_ids.keys():
self._read_inventory_data(seed_id)
# if seed id is not found read all inventories and try to find it there
if not seed_id in self.seed_ids.keys():
if self.verbosity:
print('No data found for seed id {}. Trying to find it in all known inventories...'.format(seed_id))
self.read_all()
for inv_fname, metadata_dict in self.inventory_files.items():
# use get_coordinates to check for seed_id
try:
metadata_dict['data'].get_coordinates(seed_id, time)
self.seed_ids[seed_id] = inv_fname
if self.verbosity:
print('Found metadata for station {}!'.format(seed_id))
return metadata_dict
except Exception as e:
continue
print('Could not find metadata for station {}'.format(seed_id))
return None
fname = self.seed_ids[seed_id]
return self.inventory_files[fname]
def read_all(self):
"""
Read all metadata files found in all inventories
"""
# iterate over all inventory folders
for inventory in self.inventories:
# iterate over all inventory files in the current folder
for inv_fname in os.listdir(inventory):
inv_fname = os.path.join(inventory, inv_fname)
if not self.read_single_file(inv_fname):
continue
def read_single_file(self, inv_fname):
"""
Try to read a single file as Parser/Inventory and add its dictionary to inventory files if reading sudceeded.
:param inv_fname: path/filename of inventory file
:type inv_fname: str
:rtype: None
"""
# return if it was read already
if self.inventory_files.get(inv_fname, None):
return
try:
invtype, robj = self._read_metadata_file(inv_fname)
if robj is None:
return
except Exception as e:
print('Could not read file {}'.format(inv_fname))
return
self.inventory_files[inv_fname] = {'invtype': invtype,
'data': robj}
return True
def get_coordinates(self, seed_id, time=None):
"""
Get coordinates of given seed id.
:param seed_id: Seed id such as BW.WETR..HHZ (Network.Station.Location.Channel)
:type seed_id: str
:param time: Used when a station has data available at multiple time intervals
:type time: UTCDateTime
:return: dict containing position information of the station
:rtype: dict
"""
# try most recent data if no time is specified
if not time:
time = UTCDateTime()
metadata = self.get_metadata(seed_id, time)
if not metadata:
return
try:
return metadata['data'].get_coordinates(seed_id, time)
# no specific exception defined in obspy inventory
except Exception as e:
logging.warning(f'Could not get metadata for {seed_id}')
def get_all_coordinates(self):
def stat_info_from_parser(parser):
for station in parser.stations:
station_name = station[0].station_call_letters
network_name = station[0].network_code
if not station_name in self.stations_dict.keys():
st_id = '{}.{}'.format(network_name, station_name)
self.stations_dict[st_id] = {'latitude': station[0].latitude,
'longitude': station[0].longitude,
'elevation': station[0].elevation}
def stat_info_from_inventory(inventory):
for network in inventory.networks:
for station in network.stations:
station_name = station.code
network_name = network.code
if not station_name in self.stations_dict.keys():
st_id = '{}.{}'.format(network_name, station_name)
self.stations_dict[st_id] = {'latitude': station[0].latitude,
'longitude': station[0].longitude,
'elevation': station[0].elevation}
read_stat = {'xml': stat_info_from_inventory,
'dless': stat_info_from_parser}
self.read_all()
for item in self.inventory_files.values():
inventory = item['data']
invtype = item['invtype']
read_stat[invtype](inventory)
return self.stations_dict
def get_paz(self, seed_id, time):
"""
:param seed_id: Seed id such as BW.WETR..HHZ (Network.Station.Location.Channel)
:type seed_id: str
:param time: Used when a station has data available at multiple time intervals
:type time: UTCDateTime
:rtype: dict
"""
metadata = self.get_metadata(seed_id)
if not metadata:
return
if metadata['invtype'] in ['dless', 'dseed']:
return metadata['data'].get_paz(seed_id, time)
elif metadata['invtype'] in ['resp', 'xml']:
resp = metadata['data'].get_response(seed_id, time)
return resp.get_paz(seed_id)
def _read_inventory_data(self, seed_id):
for inventory in self.inventories:
if self._read_metadata_iterator(path_to_inventory=inventory, station_seed_id=seed_id):
return
def _read_metadata_iterator(self, path_to_inventory, station_seed_id):
"""
Search for metadata for a specific station iteratively.
"""
network, station, location, channel = station_seed_id.split('.')
# seach for station seed id in filenames in invetory
fnames = glob.glob(os.path.join(path_to_inventory, '*' + station_seed_id + '*'))
if not fnames:
# search for station name in filename
fnames = glob.glob(os.path.join(path_to_inventory, '*' + station + '*'))
if not fnames:
# search for network name in filename
fnames = glob.glob(os.path.join(path_to_inventory, '*' + network + '*'))
if not fnames:
if self.verbosity:
print('Could not find filenames matching station name, network name or seed id')
return
for fname in fnames:
if fname in self.inventory_files.keys():
if self.inventory_files[fname]:
# file already read
continue
invtype, robj = self._read_metadata_file(os.path.join(path_to_inventory, fname))
try:
# robj.get_coordinates(station_seed_id) # TODO: Commented out, failed with Parser, is this needed?
self.inventory_files[fname] = {'invtype': invtype,
'data': robj}
if station_seed_id in self.seed_ids.keys():
print('WARNING: Overwriting metadata for station {}'.format(station_seed_id))
self.seed_ids[station_seed_id] = fname
return True
except Exception as e:
continue
print('Could not find metadata for station_seed_id {} in path {}'.format(station_seed_id, path_to_inventory))
def _read_metadata_file(self, path_to_inventory_filename):
"""
function reading metadata files (either dataless seed, xml or resp)
:param path_to_inventory_filename:
:return: file type/ending, inventory object (Parser or Inventory)
:rtype: (str, obspy.io.xseed.Parser or obspy.core.inventory.inventory.Inventory)
"""
# functions used to read metadata for different file endings (or file types)
read_functions = {'dless': self._read_dless,
'dataless': self._read_dless,
'dseed': self._read_dless,
'xml': self._read_inventory_file,
'resp': self._read_inventory_file}
file_ending = path_to_inventory_filename.split('.')[-1]
if file_ending in read_functions.keys():
robj, exc = read_functions[file_ending](path_to_inventory_filename)
if exc is not None:
raise exc
return file_ending, robj
# in case file endings did not match the above keys, try and error
for file_type in ['dless', 'xml']:
try:
robj, exc = read_functions[file_type](path_to_inventory_filename)
if exc is None:
if self.verbosity:
print('Read file {} as {}'.format(path_to_inventory_filename, file_type))
return file_type, robj
except Exception as e:
if self.verbosity:
print('Could not read file {} as {}'.format(path_to_inventory_filename, file_type))
return None, None
@staticmethod
def _read_dless(path_to_inventory):
exc = None
try:
parser = Parser(path_to_inventory)
except Exception as exc:
parser = None
return parser, exc
@staticmethod
def _read_inventory_file(path_to_inventory):
exc = None
try:
inv = read_inventory(path_to_inventory)
except Exception as exc:
inv = None
return inv, exc
def time_from_header(header):
"""
Function takes in the second line from a .gse file and takes out the date and time from that line.
:param header: second line from .gse file
:type header: string
:return: a list of integers of form [year, month, day, hour, minute, second, microsecond]
"""
timeline = header.split(' ')
time = timeline[1].split('/') + timeline[2].split(':')
time = time[:-1] + time[-1].split('.')
return [int(t) for t in time]
def check_time(datetime):
"""
Function takes in date and time as list and validates it's values by trying to make an UTCDateTime object from it
:param datetime: list of integers [year, month, day, hour, minute, second, microsecond]
:type datetime: list
:return: returns True if Values are in supposed range, returns False otherwise
>>> check_time([1999, 1, 1, 23, 59, 59, 999000])
True
>>> check_time([1999, 1, 1, 23, 59, 60, 999000])
False
>>> check_time([1999, 1, 1, 23, 59, 59, 1000000])
False
>>> check_time([1999, 1, 1, 23, 60, 59, 999000])
False
>>> check_time([1999, 1, 1, 23, 60, 59, 999000])
False
>>> check_time([1999, 1, 1, 24, 59, 59, 999000])
False
>>> check_time([1999, 1, 31, 23, 59, 59, 999000])
True
>>> check_time([1999, 2, 30, 23, 59, 59, 999000])
False
>>> check_time([1999, 2, 29, 23, 59, 59, 999000])
False
>>> check_time([2000, 2, 29, 23, 59, 59, 999000])
True
>>> check_time([2000, 13, 29, 23, 59, 59, 999000])
False
"""
try:
UTCDateTime(*datetime)
return True
except ValueError:
return False
# TODO: change root to datapath
def get_file_list(root_dir):
"""
Function uses a directorie to get all the *.gse files from it.
:param root_dir: a directorie leading to the .gse files
:type root_dir: string
:return: returns a list of filenames (without path to them)
"""
file_list = glob.glob1(root_dir, '*.gse')
return file_list
def checks_station_second(datetime, file):
"""
Function uses the given list to check if the parameter 'second' is set to 60 by mistake
and sets the time correctly if so. Can only correct time if no date change would be necessary.
:param datetime: [year, month, day, hour, minute, second, microsecond]
:return: returns the input with the correct value for second
"""
if datetime[5] == 60:
if datetime[4] == 59:
if datetime[3] == 23:
err_msg = 'Date should be next day. ' \
'File not changed: {0}'.format(file)
raise ValueError(err_msg)
else:
datetime[3] += 1
datetime[4] = 0
datetime[5] = 0
else:
datetime[4] += 1
datetime[5] = 0
return datetime
def make_time_line(line, datetime):
"""
Function takes in the original line from a .gse file and a list of date and
time values to make a new line with corrected date and time.
:param line: second line from .gse file.
:type line: string
:param datetime: list of integers [year, month, day, hour, minute, second, microsecond]
:type datetime: list
:return: returns a string to write it into a file.
"""
ins_form = '{0:02d}:{1:02d}:{2:02d}.{3:03d}'
insertion = ins_form.format(int(datetime[3]),
int(datetime[4]),
int(datetime[5]),
int(datetime[6] * 1e-3))
newline = line[:16] + insertion + line[28:]
return newline
def evt_head_check(root_dir, out_dir=None):
"""
A function to make sure that an arbitrary number of .gse files have correct values in their header.
:param root_dir: a directory leading to the .gse files.
:type root_dir: string
:param out_dir: a directory to store the new files somwhere els.
:return: returns nothing
"""
if not out_dir:
print('WARNING files are going to be overwritten!')
inp = str(input('Continue? [y/N]'))
if not inp == 'y':
sys.exit()
filelist = get_file_list(root_dir)
nfiles = 0
for file in filelist:
infile = open(os.path.join(root_dir, file), 'r')
lines = infile.readlines()
infile.close()
datetime = time_from_header(lines[1])
if check_time(datetime):
continue
else:
nfiles += 1
datetime = checks_station_second(datetime, file)
print('writing ' + file)
# write File
lines[1] = make_time_line(lines[1], datetime)
if not out_dir:
out = open(os.path.join(root_dir, file), 'w')
out.writelines(lines)
out.close()
else:
out = open(os.path.join(out_dir, file), 'w')
out.writelines(lines)
out.close()
print(nfiles)
def read_metadata(path_to_inventory):
"""
take path_to_inventory and return either the corresponding list of files
found or the Parser object for a network dataless seed volume to prevent
read overhead for large dataless seed volumes
:param path_to_inventory:
:return: tuple containing a either list of files or `obspy.io.xseed.Parser`
object and the inventory type found
:rtype: tuple
"""
dlfile = list()
invfile = list()
respfile = list()
# possible file extensions specified here:
inv = dict(dless=dlfile, xml=invfile, resp=respfile, dseed=dlfile[:])
if os.path.isfile(path_to_inventory):
ext = os.path.splitext(path_to_inventory)[1].split('.')[1]
inv[ext] += [path_to_inventory]
else:
for ext in inv.keys():
inv[ext] += glob.glob1(path_to_inventory, '*.{0}'.format(ext))
invtype = key_for_set_value(inv)
if invtype is None:
print("Neither dataless-SEED file, inventory-xml file nor "
"RESP-file found!")
print("!!WRONG CALCULATION OF SOURCE PARAMETERS!!")
robj = None,
elif invtype == 'dless': # prevent multiple read of large dlsv
print("Reading metadata information from dataless-SEED file ...")
if len(inv[invtype]) == 1:
fullpath_inv = os.path.join(path_to_inventory, inv[invtype][0])
robj = Parser(fullpath_inv)
else:
robj = inv[invtype]
else:
print("Reading metadata information from inventory-xml file ...")
robj = read_inventory(inv[invtype])
return invtype, robj
# idea to optimize read_metadata
# def read_metadata_new(path_to_inventory):
# metadata_objects = []
# # read multiple files from directory
# if os.path.isdir(path_to_inventory):
# fnames = os.listdir(path_to_inventory)
# # read single file
# elif os.path.isfile(path_to_inventory):
# fnames = [path_to_inventory]
# else:
# print("Neither dataless-SEED file, inventory-xml file nor "
# "RESP-file found!")
# print("!!WRONG CALCULATION OF SOURCE PARAMETERS!!")
# fnames = []
#
# for fname in fnames:
# path_to_inventory_filename = os.path.join(path_to_inventory, fname)
# try:
# ftype, robj = read_metadata_file(path_to_inventory_filename)
# metadata_objects.append((ftype, robj))
# except Exception as e:
# print('Could not read metadata file {} '
# 'because of the following Exception: {}'.format(path_to_inventory_filename, e))
# return metadata_objects
def restitute_trace(input_tuple):
def no_metadata(tr, seed_id):
print('no metadata file found '
'for trace {0}'.format(seed_id))
return tr, True
tr, metadata, unit, force = input_tuple
remove_trace = False
seed_id = tr.get_id()
mdata = metadata.get_metadata(seed_id, time=tr.stats.starttime)
if not mdata:
return no_metadata(tr, seed_id)
invtype = mdata['invtype']
inobj = mdata['data']
# check, whether this trace has already been corrected
if 'processing' in tr.stats.keys() \
and np.any(['remove' in p for p in tr.stats.processing]) \
and not force:
print("Trace {0} has already been corrected!".format(seed_id))
return tr, False
stime = tr.stats.starttime
prefilt = get_prefilt(tr)
if invtype == 'resp':
fresp = find_in_list(inobj, seed_id)
if not fresp:
return no_metadata(tr, seed_id)
fname = fresp
seedresp = dict(filename=fname,
date=stime,
units=unit)
kwargs = dict(paz_remove=None, pre_filt=prefilt, seedresp=seedresp)
elif invtype == 'dless':
if type(inobj) is list:
fname = Parser(find_in_list(inobj, seed_id))
else:
fname = inobj
paz = fname.get_paz(tr.id, datetime=tr.stats.starttime)
kwargs = dict(pre_filt=prefilt, paz_remove=paz, remove_sensitivity=True)
elif invtype == 'xml':
invlist = inobj
if len(invlist) > 1:
inventory = find_in_list(invlist, seed_id)
else:
inventory = invlist[0]
elif invtype is None:
return no_metadata(tr, seed_id)
else:
remove_trace = True
# apply restitution to data
print("Correcting instrument at station %s, channel %s" \
% (tr.stats.station, tr.stats.channel))
try:
if invtype in ['resp', 'dless']:
try:
tr.simulate(**kwargs)
print("Done")
except ValueError as e:
vmsg = '{0}'.format(e)
print(vmsg)
else:
try:
tr.attach_response(inventory)
tr.remove_response(output=unit,
pre_filt=prefilt)
except UnboundLocalError as e:
vmsg = '{0}'.format(e)
print(vmsg)
except ValueError as e:
msg0 = 'Response for {0} not found in Parser'.format(seed_id)
msg1 = 'evalresp failed to calculate response'
if msg0 not in e.message or msg1 not in e.message:
raise
else:
# restitution done to copies of data thus deleting traces
# that failed should not be a problem
remove_trace = True
return tr, remove_trace
def restitute_data(data, metadata, unit='VEL', force=False, ncores=0):
"""
takes a data stream and a path_to_inventory and returns the corrected
waveform data stream
:param data: seismic data stream
:param unit: unit to correct for (default: 'VEL')
:param force: force restitution for already corrected traces (default:
False)
:return: corrected data stream
"""
# data = remove_underscores(data)
# loop over traces
input_tuples = []
for tr in data:
input_tuples.append((tr, metadata, unit, force))
data.remove(tr)
pool = gen_Pool(ncores)
result = pool.imap_unordered(restitute_trace, input_tuples)
pool.close()
for tr, remove_trace in result:
if not remove_trace:
data.traces.append(tr)
# check if ALL traces could be restituted, take care of large datasets
# better try restitution for smaller subsets of data (e.g. station by
# station)
return data
def get_prefilt(trace, tlow=(0.5, 0.9), thi=(5., 2.), verbosity=0):
"""
takes a `obspy.core.stream.Trace` object, taper parameters tlow and thi and
returns the pre-filtering corner frequencies for the cosine taper for
further processing
:param trace: seismic data trace
:type trace: `obspy.core.stream.Trace`
:param tlow: tuple or list containing the desired lower corner
frequenices for a cosine taper
:type tlow: tuple or list
:param thi: tuple or list containing the percentage values of the
Nyquist frequency for the desired upper corner frequencies of the cosine
taper
:type thi: tuple or list
:param verbosity: verbosity level
:type verbosity: int
:return: pre-filt cosine taper corner frequencies
:rtype: tuple
..example::
>>> st = read()
>>> get_prefilt(st[0])
(0.5, 0.9, 47.5, 49.0)
"""
if verbosity:
print("Calculating pre-filter values for %s, %s ..." % (
trace.stats.station, trace.stats.channel))
# get corner frequencies for pre-filtering
fny = trace.stats.sampling_rate / 2
fc21 = fny - (fny * thi[0] / 100.)
fc22 = fny - (fny * thi[1] / 100.)
return tlow[0], tlow[1], fc21, fc22
if __name__ == "__main__":
import doctest
doctest.testmod()