Changelog

Follow up on the latest improvements and updates.

RSS

Fixed
  • Fixed Quick Compare loading error.
Improved
  • Workflow submission now supports up to 500 samples (increased from 250)
  • Improved interface for workflow submission, with progress indicator showing "X of Y submitted" during batch processing, and navigation confirmation to prevent accidental page exit.
Fixed
  • Fixed various page layout and display issues.
  • Trailing whitespaces are now trimmed from analysis job names.
Fixed
  • PNG/SVG export now works correctly on distance heatmaps in custom plots
  • API now returns 400 errors instead of 500 for malformed
    ?sort
    JSON and filtering on non-filterable schema fields
New
  • For custom workflows, we now fetch, store and verify the commit SHA corresponding to GitHub/GitLab tags.
Improved
  • Custom plot titles now truncate with ellipsis when they exceed container width instead of overflowing
Fixed
  • Custom plots now provides clearer feedback when connectivity issues occur
  • Fixed exporting multiple functional results that could time out due to empty table handling
  • Fixed ability to copy old demo notebooks
New
  • $OCX_CPU_REQUEST
    variable now injected into Custom workflow environment so that job scripts can now access the number of CPUs requested by the workflow.
  • Workflow environment variables
    $HOME
    and
    $USER
    are now set to more sensible defaults.
Improved
  • Canceling Nextflow workflows with child tasks now cancels the child tasks, instead of waiting for them to finish.
New
  • Added
    /jobs/{id}/details
    endpoint to API to expose job assets, dependencies, script, image URI, and other details with appropriate permission checks
  • Added ability for organization admins to delete unused assets from the Assets page
Fixed
  • Fixed search suggestion type from being unclickable
  • Fixed display of x-axis labels in Quick Compare charts
  • Fixed trimming on repository URLs in job editor page. Excess whitespaces will now be trimmed prior to URL validation.
  • Fixed various display issues.
  • Improved error message for Custom Plots when insufficient memory is available.

new

improved

fixed

Client Library

API

`onecodex` client library v1.1.0

New
  • Added
    updated_at
    to
    Analyses
    schema
  • onecodex.models.Workflows
    (a type of
    onecodex.models.Analyses
    ) for Custom Workflows (https://docs.onecodex.com/en/collections/9673389-your-authored-workflows-on-one-codex)
  • Added
    Jobs.run()
    for running analyses including Workflows
  • onecodex.models.Assets
    for querying and creating Workflow assets
  • Analyses.await_completion()
    blocks until an analysis reaches a terminal state using adaptive polling.
  • onecodex analyses await <analysis_id>
    CLI command, with
    --timeout
    ,
    --initial-interval
    , and
    --max-interval
    options. Also available as
    onecodex await <analysis_id>
    for backward compatibility.
  • Analyses.logs()
    fetches the job run logs for an analysis, with an optional
    tail
    parameter to limit the output to the last N lines.
  • onecodex analyses logs <analysis_id>
    CLI command, with
    --tail
    to limit the number of log lines returned (defaults to 1000). Only available for custom workflow runs.
  • Jobs.create()
    and
    Jobs.update()
    for creating and updating custom jobs from the Python client.
    assets=[...]
    accepts
    Assets
    instances directly;
    dependencies=[{"job": parent_job, "output_dir": "..."}]
    accepts
    Jobs
    instances directly.
  • onecodex jobs create
    and
    onecodex jobs update <job_id>
    CLI commands.
  • onecodex download samples
    accepts
    -s
    /
    --sample
    (repeatable) to download specific samples by ID. Mutually exclusive with
    --project
    and
    --tags
    .
  • onecodex jobs run
    accepts
    --args-json
    to pass runtime arguments as a JSON object, preserving non-string types (integers, floats, booleans, arrays, objects).
    -a
    /
    --arg
    only supports string values; use
    --args-json
    when a job argument expects another type. Mutually exclusive with
    -a
    /
    --arg
    .
  • ocx.Jobs.get("...").details()
    for retrieving CPU/memory/script/etc for Custom Workflows.
Changed
  • ApiRef
    now accepts any object with a
    field_uri
    attribute (e.g. resource model instances), in addition to the existing
    $ref
    /
    $uri
    dict shapes. This lets users pass fetched resources directly into create/update calls.
  • OneCodexBase.create()
    now validates kwargs against the schema declared in
    _allowed_methods["create"]
    (when one is set), mirroring how
    update()
    already worked. Surfaces server error messages instead of failing on response parsing.
Fixed
  • plot_pcoa
    no longer flips axis orientation between runs; PCoA eigenvector signs are now pinned so the entry with the largest absolute value on each axis is positive.
  • CLI commands that take One Codex IDs (
    onecodex samples
    ,
    analyses
    ,
    classifications
    ,
    panels
    ,
    workflows
    ,
    documents download
    ,
    jobs run
    /
    update
    ,
    upload --sample-id
    ,
    download samples --sample
    ) now validate the format up front and fail with a helpful error instead of an opaque 404.
New
  • Users can now add an optional description (up to 250 characters) when launching custom workflow runs, from the "confirm-run" page. Descriptions appear in the analysis dropdown, workflow runs dashboard (searchable), and workflow run details page.
  • It is now possible to preview .html files that are outputs from workflows.
Improved
  • Added full-name tooltip on the Run Workflows sample list, allowing users to see the sample's full name if the name is long and truncated.
  • Restored spacing between buttons on the app.onecodex.com/workflows page.
Fixed
  • Updated the save button dropdown to only save when the main button is clicked, not when selecting options from the dropdown menu.
Improved
  • We've increased the number of samples for which classification results can be exported.
Improved
  • We've extended the allowed generation time of classification results exports, to prevent timeouts for large batches of samples.
Load More