Choose packages from the approved offline catalogue. The notebook environment records the Python version, package requirements, runtime settings and a SHA-256 identity. If a replacement environment fails to prepare, the previous ready environment stays in place.
ANpy governed Python
Explore deeper questions with Python in the same project.
Use notebooks for analysis with selected file inputs and approved packages. Keep the notebook revision, environment and execution record together so the work can be reviewed in context.
Save notebooks explicitly, with checks for concurrent changes. Checkpoints and restores create fixed versions; the current saved notebook remains editable. If versions conflict, preserve the local draft and choose whether to reload or reapply it. Notebook documents use nbformat 4.
Start, interrupt, restart or stop Python only after the platform checks the workspace, notebook and environment. These controls do not give the browser access to server paths, containers or administrative credentials.
A new Python session can repeat a limited set of unchanged, previously successful setup cells when the notebook revision, code fingerprint, environment and saved execution records agree. If evidence is stale or incomplete, replay stops rather than assuming the previous state survived.
Each submitted code cell keeps its source identity, notebook revision, environment, Python session, limited output and error metadata, duration and final state. The current browser does not stream output live or provide a complete DataFrame and chart display.
Know which environment produced the analysis.
Choose packages from the approved offline catalogue. The notebook environment records the Python version, package requirements, runtime settings and a SHA-256 identity. If a replacement environment fails to prepare, the previous ready environment stays in place.
Paste or upload the list of required packages into the environment configuration.
Resolve packages from the fixed approved catalogue, without arbitrary installation from the internet.
Identify the environment by its workspace, Python version, package requirements and runtime settings.
Keep notebook revisions without overwriting another change.
Save notebooks explicitly, with checks for concurrent changes. Checkpoints and restores create fixed versions; the current saved notebook remains editable. If versions conflict, preserve the local draft and choose whether to reload or reapply it. Notebook documents use nbformat 4.
Control a session within its authorised workspace.
Start, interrupt, restart or stop Python only after the platform checks the workspace, notebook and environment. These controls do not give the browser access to server paths, containers or administrative credentials.
Resume only from setup that can be verified.
A new Python session can repeat a limited set of unchanged, previously successful setup cells when the notebook revision, code fingerprint, environment and saved execution records agree. If evidence is stale or incomplete, replay stops rather than assuming the previous state survived.
Stopping or restarting does not preserve Python process memory.
Eligible, unchanged earlier cells may be repeated in a new session.
Changed code, a changed environment or missing records from earlier cells prevent replay rather than guessing which objects exist.
Review the code and context behind a cell's result.
Each submitted code cell keeps its source identity, notebook revision, environment, Python session, limited output and error metadata, duration and final state. The current browser does not stream output live or provide a complete DataFrame and chart display.
Choose exactly which file version the notebook can read.
Select an authorised Drive file or shared file, bind its version and prepare it as a read-only notebook input. A configured runtime is required. This selection does not grant access to every project file or to external systems.
Next step
Start with the flow that cannot afford ambiguity.
Bring the systems, owners, rules, delivery obligations and evidence requirements that matter.
Discuss a critical flowIn brief
What does this part of advanexus do?
Use notebooks for analysis with selected file inputs and approved packages. Keep the notebook revision, environment and execution record together so the work can be reviewed in context.
What should I check before using it?
Drive mounting, direct table-to-DataFrame access and full collaborative notebook behaviour are not implicit capabilities.