Comment on results of an experiment#

Comments serve experiment tracking greatly since they allow record thoughts on experiment results or expectations before them. They are more flexible, expressive, and less formal than parameters for example. They also are a basis for collaboration since multiple people can comment in a shared repo.

Add a Comment#

In Cascade you can have comments in any Traceable object. Comments are recorded in meta. In case of disk-based Traceables they are automatically synced to disk after creation.

from cascade.models import ModelLine

line = ModelLine("line")
model = line.create_model()

Notice how you can comment not only on models. Models will save comments to disk only when saved, when lines and repos will sync meta with disk automatically.

model.comment("This experiment will be the best")
line.comment("Did you know that you can comment on any Traceable?")

Remove a Comment#

Every comment has an integer ID starting with 1, using which you can locate and delete the comment.

model.remove_comment(1)

Using Cascade CLI#

CLI is the best way to use this feature.

Locate to the repo, line or model folder and then use:

cascade comment add

You will be prompted to enter a comment afterwards. Hit enter when you are ready.

You can display all the comments using:

cascade comment ls

And delete using:

cascade comment rm 1