DASCore

A python library for distributed fiber optic sensing.

Code

Documentation [stable, development]

Citation

Chambers, D., Jin, G., Tourei, A., Issah, A. H. S., Lellouch, A., Martin, E., Zhu, D., Girard, A., Yuan, S., Cullison, T., Snyder, T., Kim, S., Danes, N., Pnithan, N., Boltz, M. S. & Mendoza, M. M. (2024). DASCore: a Python Library for Distributed Fiber Optic Sensing. Seismica, 3(2).

Version-specific documentation builds are attached to the release pages.

Highlights

DASCore is a Python library for distributed acoustic sensing (DAS). It provides:

1. IO support for various DAS data formats
2. Common processing routines
3. Basic visualizations

Note

DASCore is a foundational package of the DAS Data Analysis Ecosystem (DASDAE). It facilitates rapid development of other DASDAE packages that do more specialized analysis/visualization.

Supported file formats

name	version	scan	get_format	read	write
APSENSING	10	✅	✅	✅	❌
DASDAE	1	✅	✅	✅	✅
DASHDF5	1.0	✅	✅	✅	❌
FEBUS	2	✅	✅	✅	❌
FEBUS	1	✅	✅	✅	❌
H5SIMPLE	1	✅	✅	✅	❌
NEUBREXRFS	1	✅	✅	✅	❌
NEUBREXDAS	1	✅	✅	✅	❌
OPTODAS	8	✅	✅	✅	❌
PICKLE		❌	✅	✅	✅
PRODML	2.0	✅	✅	✅	❌
PRODML	2.1	✅	✅	✅	❌
RSF	1	❌	❌	❌	✅
SEGY	2	✅	✅	✅	❌
SENTEK	5	✅	✅	✅	❌
SILIXA_H5	1	✅	✅	✅	❌
TDMS	4713	✅	✅	✅	❌
TERRA15	4	✅	✅	✅	❌
TERRA15	5	✅	✅	✅	❌
TERRA15	6	✅	✅	✅	❌
WAV		❌	❌	❌	✅
XMLBINARY	1	✅	✅	✅	❌

Introductory usage

Read a file

import dascore as dc
from dascore.utils.downloader import fetch

# get a path to an example file, replace with your path
file_path = fetch('terra15_das_1_trimmed.hdf5')

spool = dc.spool(file_path)
patch = spool[0]

Working with a directory of DAS files

import dascore as dc
from dascore.utils.downloader import fetch

# get a path to a directory of das files, replace with your path
directory_path = fetch('terra15_das_1_trimmed.hdf5').parent

spool = (
    # Create a spool to interact with directory data
    dc.spool(directory_path)
    # Index the directory contents
    .update()
    # Sub-select a specific time range
    .select(time_min=('2020-01-01', ...))
    # Specify chunk of the output patches
    .chunk(time=60, overlap=10)
)

Get patches (array w/ metadata)

import dascore as dc

spool = dc.get_example_spool('diverse_das')

# Spools are like lists, you can get patches through iteration
for patch in spool:
    ...

# or through indexing
patch = spool[0]

Perform processing

import dascore as dc

patch = dc.get_example_patch('random_das')

out = (
    # Decimate along time axis (keep every 8th sample)
    patch.decimate(time=8)
    # Detrend along the distance dimension
    .detrend(dim='distance') 
    # Apply 10Hz low-pass filter along time dimension
    .pass_filter(time=(..., 10))
)

Visualize

import dascore as dc

patch = dc.get_example_patch('example_event_2')

patch.viz.waterfall(show=True, scale=0.2);

Installation

Use pip or conda to install DASCore:

pip install dascore

conda install dascore -c conda-forge

See this recipe for installation with Docker.

See the contributing docs for development installations.

Feedback and Support

Use the project discussions to ask a question.

Use the project issues to report an issue.