search

vc1492a / tidd

An approach for detecting tsunamis using anomaly detection anomalies on sTec d/dt data from orbiting GPS satellites.
Other
5 stars 1 forks source link

Determine a name for the library that facilitate experiments #18

Closed vc1492a closed 3 years ago

vc1492a commented 4 years ago

Thinking forward, the library which facilitates experiments should have a distinct name that:

This should be completed prior to adding installation instructions in the readme.

vc1492a commented 4 years ago

@hamlinliu17 @MichelaRavanelli I have included the functionality in the repository (specifically, the feature/library_basics branch to allow for the core functions (reading data, future data transformations, modeling experiment runs, etc.) to be imported and used as an installable library. Right now, the library is generically called src, but it would perhaps begin thinking about what this library could be called as we continue working in these early stages.

vc1492a commented 3 years ago

We may also update the repository name if we decide on a name for the library.

vc1492a commented 3 years ago

Toolkit for Ionospheric Disturbance Detection (TIDD)? Thoughts @hamlinliu17 and @MichelaRavanelli ? Trying to keep it general.

Looks like tidd is available on the Python package index also.

hamlinliu17 commented 3 years ago

@vc1492a That sounds good with me. Will the library mostly contain the functionality to read data and encode the data into images?

vc1492a commented 3 years ago

The library will include the functionality needed to read the data, convert it to images, generate a labeled and unlabeled dataset for model training, train the model, and test its performance, reporting etc. The point is to modularize each component so that it can be developed further into the future.

For example, future versions of the toolkit should maintain the reproducibility of existing / prior experiments while also extending the functionality as current work dictates.

vc1492a commented 3 years ago

Along with that, we should do some diagramming and determine what the best structure for the code will be - I like the idea of a layered interface. E.g., the data read and data conversion bits are all part of a pipe (for a specific ground station + satellite), and an experiment can contain multiple pipes (I'm thinking out loud right now).

MichelaRavanelli commented 3 years ago

@vc1492a I like that name! Sounds good to me!

vc1492a commented 3 years ago

TIDD it is! I'll create a new branch for this issue soon that will update the repository for the name change.

vc1492a commented 3 years ago

Changed the tag from in progress to review since the only thing to do is to review the change and make the change once we stabilize the current changes on main.

vc1492a commented 3 years ago

Went ahead and changed the name of the library to tidd on Github - @hamlinliu17 be sure to update the remote on your end.