utils

Back to API tree index

Purpose

Cross-cutting utility functions and run/path helpers.

Module summary

No module-level description available.

Top-level functions

Function	Summary
A_lm_halfinclin	No docstring summary available.
A_lm_inclin	No docstring summary available.
amend_indices	Joins blocks together if they are separated by less than n_tol
asdf_fromfile
band_filter	Creates desired filter, and computes its impulse response length
before_write_save_old	No docstring summary available.
bincent	No docstring summary available.
bool2int	No docstring summary available.
change_filter_times_fd	Converts a conditioned filter to a different time array by zero-padding in time-domain
change_filter_times_td	Converts a conditioned filter to a different time array by zero-padding in time-domain
change_wf_fd_grid
checkempty	No docstring summary available.
close_hdf5
coherent_score	Combine H1 terms from scores_vetoed_max, returns scalar for single element
colorbar	No docstring summary available.
condition_filter	Compute support, weight, and truncate input frequency domain filter
condition_filter_td	Compute support, weight, and truncate input time domain filter
create_chirp_mass_directory_dict	Reads the names of all subdirectories, and creates a dictionary of dictionaries, with the outer dictionary having chirp mass ids as keys and the inner dictionary having subbank ids as keys and subdirectory names as values
define_coarser_mask
delete_hdf5_datasets	Convenience function to prune some leaves from a hdf5 file
env_init_lines	Creates text snippet to be added before submitting jobs to clusters
extract_filename	Extracts the filename from a string representation of a buffer object
extract_parts	Extracts the parts of the directory name based on the prefix and suffix
find_closest_coarse_calphas	Finds the closest associated coarse calphas to given fine calphas
gen_step_fd	No docstring summary available.
get_coincident_json_filelist	Assumes that the file names go like “Det-Detn-….” where Detn is the key
get_detector_fnames
get_dirs	Gives a list of length n_runs with each entry being a dictionary of dictionaries with the outer dictionary having chirp mass ids as keys and the inner dictionary having subbank ids as keys and subdirectory names as values. Figures out the number of subbanks and chirp mass ids by itself
get_dtype	No docstring summary available.
get_evname_from_tgps	No docstring summary available.
get_gwf_channel_names	No docstring summary available.
get_hdf5_file	Returns a hdf5 file object with the given mode, given a source. Warning, if the file is already open in read-only mode, we can’t return a writeable version.
get_HL_filenames	No docstring summary available.
get_injection_details	No docstring summary available.
get_json_fname	TODO: What if we choose a different fmax? Assumes that the file names go like “Det-Detn-….” where Detn is the detector key
get_left_right_fnames	No docstring summary available.
get_lsc_event_times	No docstring summary available.
get_O3_lsc_pe_samples	Returns a PESummary object to interact with the LSC PE samples
get_root_dirs	Gets root directories by run, used since O3a is on scratch
get_run	Returns the run name for a given GPS time
get_spectrum	Function used to dechirp Gets a trigger in the format of trigger, bank_id
get_strain_fnames	Returns the strain file names given a time
get_tgps_from_evname	No docstring summary available.
handle_missing_run	No docstring summary available.
hilbert_transform	No docstring summary available.
hole_edges	No docstring summary available.
hole_edges_to_mask	No docstring summary available.
incoherent_score
index_after_removal	No docstring summary available.
index_limits	Function that decides which indices to include in the average such that we always average window_size indices to avoid \`regression-to-mean’ artifacts due to fewer samples near holes
interpolate_asd	get log-log interpolant of ASD with old_f[0] >= 0,
interpolate_wf_fd	get log-log interpolant of amp & lin of phase with old_f[0] >= 0,
is_close_to
is_in_run	Check if a GPS time is in a given run, works with HM/non-standard runs
is_LIGO_valid	No docstring summary available.
is_LIGO_valid_between	Take two gps times t1 and t2 and return: 0 if the hole chunk from t1 to t2 is invalid 1 if it is partially valid 2 if it is fully valid where valid means that both H and L have all the [‘DATA’, ‘CBC_CAT1’, ‘CBC_CAT2’, ‘CBC_CAT3’] flags True. (e.g. for assessing whether a waveform between t1 and t2 had the right to be found or not)
is_within	No docstring summary available.
load_dict_from_hdf5_attrs	Loads a dictionary from the attributes of a hdf5 file
load_pesummary_io	I need to do some monkey patching to make it work on my machine
m1_m2_mt	complete m1, m2, mt or check consistency if all given –> return m1, m2, mt
m1_m2_s1_s2_to_chieff_chia	No docstring summary available.
make_LIGO_mask	Run this only once to make the file /data/bzackay/GW/LIGO_holes.npy
make_template_ids	No docstring summary available.
make_trigger_ids	No docstring summary available.
mass_conversion	Computes all conversions given a subset of parameters describing the masses
match	Computes match, or cosine, between waveforms
mmap_h5	Memory mapping is much faster for single row access than non-chunked hdf5, as long as we have a 64 bit architecture It is slower for random access, as the kernel version on the IAS systems is older and doesn’t activate the madvise system call WARNING: If the hdf5 file is already open (e.g., in a writeable manner), the values in the mmap object might not be updated until the changes to the hdf5 file are flushed (despite the description of the dedault UNIX driver, it has a small buffer, set by rdcc_nbytes)!
multiprocessing	A shorthand function to apply multiprocessing in a one liner
notch_filter	Creates notch filter
notch_filter_sos	Defines set of sos filters to apply to notch out lines
offset_background	Finds the amount to shift the detectors’ data streams by. It returns an integer multiple of dt_shift
orthogonalize_split	Function to compute orthogonalized scores to those in the L subset
pars_from_pars	WARNING: this sees keys ‘chi1’ and ‘chi2’ as being the vector spins [sjx, sjy, sjz] \*UNLIKE\* in other places where chij = sqrt(sjx^2 + sjy^2 + sjz^2) complete intrinsic parameter dictionary from sufficient parts
plot_veto_details	Make four-panel diagnostic plot for vetoes
populate_magic_methods	Decorator to populate magic methods for HDF5DatasetSubset, they enable us to interact with it like it’s a numpy array when needed
preprocess_wildcards	No docstring summary available.
q_and_eta	No docstring summary available.
read_hdf5_node	Convenience function to read/create a group from a hdf5 file
remove_bad_times	Removes elements of dat that are in the same bucket as in bad_time_list
remove_old_versions	No docstring summary available.
rm_suffix	Utility to change the extension of a path
s1z_s2z_chieff_chia_from_pars	\*NOTE\* this uses chia = (m1\*s1z - m2\*s2z) / (m1 + m2)
safe_concatenate	No docstring summary available.
safelen	No docstring summary available.
save_dict_to_hdf5_attrs	Saves a dictionary to the attributes of a hdf5 file
scalar	No docstring summary available.
sigma_from_median	Computes sigma from median for an array with Gaussian samps + outliers Silently returns 1 if we passed in an empty array
sinc_interp_by_factor_of_2
sinc_interp_x2D	exact same as sinc_interp_by_factor_of_2() except along last axis of two dimensional array x2D
sine_gaussian	Function to return time-domain sine-gaussian pulses (ready for FFT)
splitarray	Splits a single or multi-dimensional array into sub-arrays based on a coordinate
standardize_run_name	Maintain consistency in run names, mainly used to confirm with the GWOSC naming conventions
submask
threshold_rv
track_job	Track a multiprocessing job that was submitted in chunks
unbias_split	Function to compute split scores - expectations from total score
write_hdf5_node	Convenience function to create/overwrite a leaf in a hdf5 file
Y_lm	NOTE: azim = pi/2 - vphi
Y_lm_halfinclin	NOTE: azim = pi/2 - vphi

Classes

Class	Summary
CustomDefaultdict	No class docstring summary available.
CustomHDF5File	Custom h5py file object that wraps datasets with EditableHDF5Dataset on access
CustomHDF5Group	Custom h5py group object that wraps datasets with EditableHDF5Dataset on access
EditableHDF5Dataset	Wrapper of hdf5 dataset that returns references instead of copies when indexed if it is multidimensional or variable-length
HDF5DatasetSubset	Simulates a copy-on-write reference to a subset of a multidimensional or variable-length HDF5 dataset. This is needed because by default, edits to a HDF5 dataset’s elements don’t propagate back to the underlying dataset
NumpyEncoder	No class docstring summary available.
TupleEncoder	No class docstring summary available.