Export via PachCTL

To export your data with PachCTL:

  1. List the files in the given directory:

    pachctl list file <repo>@<branch>:<dir>

    Example:

    pachctl list myrepo@master:labresults
    NAME                                                    TYPE SIZE
    /labresults/T1606331395-LIPID-PATID2-CLIA24D9871327.txt file 101B
    /labresults/T1606707557-LIPID-PATID1-CLIA24D9871327.txt file 101B
    /labresults/T1606707579-LIPID-PATID3-CLIA24D9871327.txt file 100B
    /labresults/T1606707597-LIPID-PATID4-CLIA24D9871327.txt file 101B
  2. Get the contents of a specific file:

    pachctl get file <repo>@<branch>:<path/to/file>
    pachctl get file myrepo@master:/labresults/T1606331395-LIPID-PATID2-CLIA24D9871327.txt
    PID|PATID2
    ORC|ORD777889
    OBX|1|NM|2093-3^Cholesterol|212|mg/dL
    OBX|2|NM|2571-8^Triglyceride|110|mg/dL
    Note

    You can view the parent, grandparent, and any previous commit by using the caret (^) symbol followed by a number that corresponds to an ancestor in sequence:

    • View a parent commit

      pachctl list commit <repo>@<branch-or-commit>^:<path/to/file>
      pachctl get file <repo>@<branch-or-commit>^:<path/to/file>
    • View an <n> parent of a commit

      pachctl list commit <repo>@<branch-or-commit>^<n>:<path/to/file>
      pachctl get file <repo>@<branch-or-commit>^<n>:<path/to/file>

      Example:

      pachctl get file datas@master^4:user_data.csv

    If the file does not exist in that revision, HPE Machine Learning Data Management displays an error message.