Dlisio

dlisio - DLIS/LIS File Reader

Quick Reference

import dlisio

# Open DLIS file (returns generator of logical files)
with dlisio.dlis.load('well.dlis') as (f, *rest):
    frame = f.frames[0]
    curves = frame.curves()

    # Access by channel name
    depth = curves['DEPTH']
    gr = curves['GR']

    # File metadata
    for origin in f.origins:
        print(origin.well_name, origin.field_name)

Key Classes

| Class | Purpose | |-------|---------| | PhysicalFile | Container returned by dlis.load() | | LogicalFile | Independent dataset within physical file | | Frame | Group of channels with common sampling | | Channel | Individual log curve with metadata | | Origin | Well and file metadata |

Essential Operations

Read Curves to DataFrame

import pandas as pd

with dlisio.dlis.load('well.dlis') as (f, *_):
    frame = f.frames[0]
    curves = frame.curves()
    df = pd.DataFrame(curves)
    df.set_index('DEPTH', inplace=True)

Access Channel and Origin Metadata

with dlisio.dlis.load('well.dlis') as (f, *_):
    # Origin metadata
    for origin in f.origins:
        print(f"Well: {origin.well_name}, Field: {origin.field_name}")

    # Channel properties
    for ch in f.frames[0].channels:
        print(f"{ch.name}: {ch.units}, dim={ch.dimension}")

Find Channels Across Frames

with dlisio.dlis.load('well.dlis') as (f, *_):
    # By exact name or regex
    channels = f.find('CHANNEL', '.*GR.*', regex=True)

    # Find frame containing specific channel
    for frame in f.frames:
        if 'GR' in [ch.name for ch in frame.channels]:
            curves = frame.curves()
            break

Handle Array Channels

with dlisio.dlis.load('well.dlis') as (f, *_):
    curves = f.frames[0].curves()
    for name, data in curves.items():
        if data.ndim > 1:
            print(f"{name}: shape = {data.shape}")  # Image/waveform

Common Object Types

| Object Type | Description | |-------------|-------------| | ORIGIN | File/well metadata | | FRAME | Channel grouping with index | | CHANNEL | Log curve definition | | TOOL | Logging tool info | | PARAMETER | Constants and settings |

Common Curve Names

| Curve | Description | |-------|-------------| | DEPT, DEPTH, TDEP | Depth curves | | GR | Gamma ray | | NPHI | Neutron porosity | | RHOB | Bulk density | | DT, DTC | Compressional slowness | | RT, ILD | Resistivity |

Error Handling

dlisio.dlis.set_encodings(['utf-8', 'latin-1'])

try:
    with dlisio.dlis.load('file.dlis') as files:
        for f in files:
            curves = f.frames[0].curves()
except Exception as e:
    print(f"Error: {e}")

DLIS vs LAS Comparison

| Feature | DLIS | LAS | |---------|------|-----| | Format | Binary | ASCII | | Multi-frame | Yes | No | | Array data | Yes | Limited | | Metadata | Rich | Basic |

When to Use vs Alternatives

| Tool | Best For | |------|----------| | dlisio | Reading DLIS/RP66 binary files, multi-frame data, image logs | | lasio | LAS (ASCII) well log files, simpler format, widely supported | | welly | Higher-level well data management, curve processing, projects |

Use dlisio when your data is in DLIS (RP66) format. DLIS files are common from modern logging tools and contain multi-frame, array, and image data that LAS cannot represent.

Use lasio instead when your data is in LAS format. LAS is ASCII-based, simpler, and more widely supported. Convert DLIS to LAS when downstream tools require it.

Use welly instead when you need well-level data management with curve processing, formation tops, and multi-well projects after initial file loading.

Common Workflows

Read and convert DLIS to DataFrame

- [ ] Load file with `dlisio.dlis.load()`, handle encoding if needed
- [ ] List logical files and frames to understand file structure
- [ ] Inspect channels: names, units, dimensions per frame
- [ ] Extract curves from target frame with `frame.curves()`
- [ ] Handle array/image channels separately (ndim > 1)
- [ ] Convert scalar curves to DataFrame with `pd.DataFrame(curves)`
- [ ] Export to CSV or convert to LAS format

References

• [DLIS File Structure](references/dlis_structure.md) - RP66 format specification
• [Frames and Channels](references/frame_channels.md) - Working with frames and channels

Scripts

• [scripts/dlistolas.py](scripts/dlistolas.py) - Convert DLIS to LAS format

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

• Author: SteadfastAsArt
• Source: SteadfastAsArt/geoscience-skills
• License: MIT

Install and usage instructions live in the source repository linked above.