ariewjy / lasio

Python library for reading and writing well data using Log ASCII Standard (LAS) files

lasio

Documentation (stable) • Documentation (main branch)

Read and write Log ASCII Standard files with Python.

This is a Python 3.7+ package to read and write Log ASCII Standard (LAS) files, used for borehole data such as geophysical, geological, or petrophysical logs. It's compatible with versions 1.2 and 2.0 of the LAS file specification, published by the Canadian Well Logging Society. Support for LAS 3 is being worked on.

lasio is primarily for reading and writing data and metadata to and from LAS files. It is designed to read as many LAS files as possible, including those containing common errors and non-compliant formatting. It can be used directly, but you may want to consider using some other packages, depending on your priorities:

• welly is a Python package that uses lasio for I/O but provides a lot more functionality aimed at working with curves, wells, and projects. I would recommend starting there in most cases, to avoid re-inventing the wheel!
• lascheck is focused on checking whether your LAS file meets the specifications.
• lasr is an R package which is designed to read large amounts of data quickly from LAS files; this is a great thing to check out if speed is a priority for you, as lasio is not particularly fast.
• LiDAR surveys are also called "LAS files", but they are quite different and lasio will not help you -- check out laspy instead.

lasio stopped supporting Python 2.7 in August 2020. The final version of lasio with Python 2.7 support is version 0.26.

Code of conduct

See our code of conduct.

Quick start

For the minimum working requirements, you'll need numpy installed. Install lasio with:

$ pip install lasio

To make sure you have everything, use this to ensure pandas, chardet, and openpyxl are also installed:

$ pip install lasio[all]

Example session:

>>> import lasio

You can read the file using a filename, file-like object, or URL:

>>> las = lasio.read("sample_rev.las")

Data is accessible both directly as numpy arrays

>>> las.keys()
['DEPT', 'DT', 'RHOB', 'NPHI', 'SFLU', 'SFLA', 'ILM', 'ILD']
>>> las['SFLU']
array([ 123.45,  123.45,  123.45, ...,  123.45,  123.45,  123.45])
>>> las['DEPT']
array([ 1670.   ,  1669.875,  1669.75 , ...,  1669.75 ,  1670.   ,
        1669.875])

and as CurveItem objects with associated metadata:

>>> las.curves
[CurveItem(mnemonic=DEPT, unit=M, value=, descr=1  DEPTH, original_mnemonic=DEPT, data.shape=(29897,)),
CurveItem(mnemonic=DT, unit=US/M, value=, descr=2  SONIC TRANSIT TIME, original_mnemonic=DT, data.shape=(29897,)),
CurveItem(mnemonic=RHOB, unit=K/M3, value=, descr=3  BULK DENSITY, original_mnemonic=RHOB, data.shape=(29897,)),
CurveItem(mnemonic=NPHI, unit=V/V, value=, descr=4   NEUTRON POROSITY, original_mnemonic=NPHI, data.shape=(29897,)),
CurveItem(mnemonic=SFLU, unit=OHMM, value=, descr=5  RXO RESISTIVITY, original_mnemonic=SFLU, data.shape=(29897,)),
CurveItem(mnemonic=SFLA, unit=OHMM, value=, descr=6  SHALLOW RESISTIVITY, original_mnemonic=SFLA, data.shape=(29897,)),
CurveItem(mnemonic=ILM, unit=OHMM, value=, descr=7  MEDIUM RESISTIVITY, original_mnemonic=ILM, data.shape=(29897,)),
CurveItem(mnemonic=ILD, unit=OHMM, value=, descr=8  DEEP RESISTIVITY, original_mnemonic=ILD, data.shape=(29897,))]

Header information is parsed into simple HeaderItem objects, and stored in a dictionary for each section of the header:

>>> las.version
[HeaderItem(mnemonic=VERS, unit=, value=1.2, descr=CWLS LOG ASCII STANDARD -VERSION 1.2, original_mnemonic=VERS),
HeaderItem(mnemonic=WRAP, unit=, value=NO, descr=ONE LINE PER DEPTH STEP, original_mnemonic=WRAP)]
>>> las.well
[HeaderItem(mnemonic=STRT, unit=M, value=1670.0, descr=, original_mnemonic=STRT),
HeaderItem(mnemonic=STOP, unit=M, value=1660.0, descr=, original_mnemonic=STOP),
HeaderItem(mnemonic=STEP, unit=M, value=-0.125, descr=, original_mnemonic=STEP),
HeaderItem(mnemonic=NULL, unit=, value=-999.25, descr=, original_mnemonic=NULL),
HeaderItem(mnemonic=COMP, unit=, value=ANY OIL COMPANY LTD., descr=COMPANY, original_mnemonic=COMP),
HeaderItem(mnemonic=WELL, unit=, value=ANY ET AL OIL WELL #12, descr=WELL, original_mnemonic=WELL),
HeaderItem(mnemonic=FLD, unit=, value=EDAM, descr=FIELD, original_mnemonic=FLD),
HeaderItem(mnemonic=LOC, unit=, value=A9-16-49, descr=LOCATION, original_mnemonic=LOC),
HeaderItem(mnemonic=PROV, unit=, value=SASKATCHEWAN, descr=PROVINCE, original_mnemonic=PROV),
HeaderItem(mnemonic=SRVC, unit=, value=ANY LOGGING COMPANY LTD., descr=SERVICE COMPANY, original_mnemonic=SRVC),
HeaderItem(mnemonic=DATE, unit=, value=25-DEC-1988, descr=LOG DATE, original_mnemonic=DATE),
HeaderItem(mnemonic=UWI, unit=, value=100091604920, descr=UNIQUE WELL ID, original_mnemonic=UWI)]
>>> las.params
[HeaderItem(mnemonic=BHT, unit=DEGC, value=35.5, descr=BOTTOM HOLE TEMPERATURE, original_mnemonic=BHT),
HeaderItem(mnemonic=BS, unit=MM, value=200.0, descr=BIT SIZE, original_mnemonic=BS),
HeaderItem(mnemonic=FD, unit=K/M3, value=1000.0, descr=FLUID DENSITY, original_mnemonic=FD),
HeaderItem(mnemonic=MATR, unit=, value=0.0, descr=NEUTRON MATRIX(0=LIME,1=SAND,2=DOLO), original_mnemonic=MATR),
HeaderItem(mnemonic=MDEN, unit=, value=2710.0, descr=LOGGING MATRIX DENSITY, original_mnemonic=MDEN),
HeaderItem(mnemonic=RMF, unit=OHMM, value=0.216, descr=MUD FILTRATE RESISTIVITY, original_mnemonic=RMF),
HeaderItem(mnemonic=DFD, unit=K/M3, value=1525.0, descr=DRILL FLUID DENSITY, original_mnemonic=DFD)]

The data is stored as a 2D numpy array:

>>> las.data
array([[ 1670.   ,   123.45 ,  2550.   , ...,   123.45 ,   110.2  ,   105.6  ],
       [ 1669.875,   123.45 ,  2550.   , ...,   123.45 ,   110.2  ,   105.6  ],
       [ 1669.75 ,   123.45 ,  2550.   , ...,   123.45 ,   110.2  ,   105.6  ],
       ...,
       [ 1669.75 ,   123.45 ,  2550.   , ...,   123.45 ,   110.2  ,   105.6  ],
       [ 1670.   ,   123.45 ,  2550.   , ...,   123.45 ,   110.2  ,   105.6  ],
       [ 1669.875,   123.45 ,  2550.   , ...,   123.45 ,   110.2  ,   105.6  ]])

You can also retrieve and load data as a pandas DataFrame, build LAS files from scratch, write them back to disc, and export to Excel, amongst other things.

See the package documentation for more details.

Contributing

Contributions are invited and welcome.

See Contributing for how to get started.

License

MIT