Skip to content
This repository has been archived by the owner on Aug 29, 2023. It is now read-only.

Representative default variables #590

Closed
forman opened this issue Mar 26, 2018 · 2 comments
Closed

Representative default variables #590

forman opened this issue Mar 26, 2018 · 2 comments
Assignees

Comments

@forman
Copy link
Member

forman commented Mar 26, 2018

Expected behavior

When a dataset is opened in the Cate Desktop, the geo-spatial variable displayed in the World View should be a representative default variable.

Actual behavior

Cate Desktop displays the first geo-spatial variable it finds in a dataset. This is often not what a user expects, for example the if the first variable contains quality flags or is a statistical by-product rather then being a geo-spatial and geo-physical variable that represents the ECV of a given dataset.

Specifications

Cate Desktop 1.0 - 2.0.0-dev.8

@forman forman self-assigned this Mar 26, 2018
@papesci
Copy link
Contributor

papesci commented Mar 27, 2018

The task to find the most significant variable for each data set i think can be done only by a human. We can generate maps of all the variables to help the selection process but the final decision should be done by a human.

We can apply some statistics method like P-value or even a more sophisticated Random forest to identify the principal component but what is important for a machine might be no so important for humans and we risk to show a plot of 'number of observation' variable that i think is not so interested.

@forman
Copy link
Member Author

forman commented Mar 28, 2018

Oh, I should have said that default variables should be provided by configuration, that is, by humans.

Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.
Projects
None yet
Development

No branches or pull requests

2 participants