Download and export

CloudSprite keeps the original file you uploaded and the traces parsed from it. Pick the one that matches the next tool.

Original files

On the dataset page, download the Touchstone, CSV, or image you (or a capture) put there. That is the vendor-faithful copy: header, comments, and all ports. Use it when you need to re-open the capture in the VNA software or send a customer the file they would recognize.

Traces

Notebooks and the SDK read traces (s21.x, s21.y). Publish computed traces as a new dataset rather than “exporting a spreadsheet of SDD21” if the rest of the team should plot them in CloudSprite. If a downstream tool needs a Touchstone, write that file locally from scikit-rf after to_network(), then upload it to the result dataset if it should live in the project.

Notebook snapshots

A snapshot freezes the overlay and the parameters as they were at review time. Download or share the snapshot when the artifact is “what we signed off,” not a live plot that will move when someone re-parses a file. Live traces keep following the dataset; a snapshot does not.

What not to do

• Do not treat a published result as a download-only sidecar. It is a dataset in the same project.
• Do not script against raw file URLs — they are not a stable interface. Use the app download or the SDK helpers in the generated reference.
• Viewers can download what they can see; they cannot publish a replacement.