Skip to content

Grid Overrides #612

Closed

Description

@BrianMichell

AutoChannelWrap

We believe that this override does not require modification. Unit tests pass.

NonBinned

Input Parameters

  • List of non-binned dimensions
  • Chunk size of new dimension

Behavior

From the template, we will:

  1. Drop all dimensions contained in the list of non-binned dimensions
  2. Add a new dimension trace to the N-1 position
    • The size of this dimension will be the product of all unique elements in the list of non-binned dimensions
  3. Add one new coordinate per element in the list of non-binned dimensions
    • These coordinates will have the dimensions up to trace

Example

Template with [shot_point, cable, channel, azimuth, offset, sample] and non-binned [azimuth, offset].

The new template will have dimensions [shot_point, cable, channel, trace, sample] and new coordinates azimuth and offset with dimensions [shot_point, cable, channel, trace]

The chunk size of sample will be equal to the input parameter.

DuplicateIndex

Input Parameters

Just a flag (TBD).

Behavior

We will add a new dimension called trace to the provided template at the N-1 position. This dimension must have a chunk size of 1. Coordinates will not be modified.

ChannelWrap

We believe this can be safely removed as it is superseded by AutoChannelWrap.

CalculateCable

We believe this can be safely removed as it is superseded by AutoChannelWrap.

AutoShotwrap

We believe this can be safely removed as it is not necessary.
This is necessary, but only for a specific schema.

Activity

  1. dmitriyrepin commented on Aug 22, 2025

    @dmitriyrepin
    Contributor

    Looks good. The discussed priorities are:
    P1 - AutoChannelWrap
    P2 - NonBinned
    P3 - DuplicateIndex
    P98 - ChannelWrap, CalculateCable (no need to hurry with removal)
    P99 - AutoShotwrap - not a standard operation and will require a custom template

  2. dmitriyrepin commented on Aug 31, 2025

    @dmitriyrepin
    Contributor

    @tasansal , @BrianMichell
    For 'NonBinned':
    If the original template had the following

    • dimensions ["shot_point", "cable", "channel", "time"]
    • coordinate that is dependent on some those dimensions. e.g., "source_coord_x" -> ["shot_point", "cable", "channel"]
      what should happen to the coordinate (e.g., "source_coord_x") if declare the dimension "channel" NonBinned and remove the "channel" coordinate?
  3. markspec commented on Sep 3, 2025

    @markspec
    Contributor

    Is there a reason we are no longer indexing on shot_line and are not using AutoShotWrap. We exclusively use this when ingesting pre-stack data for ML training.

  4. markspec commented on Sep 3, 2025

    @markspec
    Contributor

    Also gun!

  5. tasansal commented on Sep 5, 2025

    @tasansal
    Contributor
  6. added theissue type on Sep 8, 2025
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Labels

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions