Skip to content

Add degradation metric computation implementation - #91

Open
Ashvin (ashvina) wants to merge 5 commits into
microsoft:mainfrom
ashvina:fix90-stability-analysis
Open

Add degradation metric computation implementation#91
Ashvin (ashvina) wants to merge 5 commits into
microsoft:mainfrom
ashvina:fix90-stability-analysis

Conversation

@ashvina

@ashvina Ashvin (ashvina) commented Jun 28, 2023

Copy link
Copy Markdown
Contributor

This change adds a basic implementation for computing degradation metric as described in the paper. The implementation, along with the helper class, can compute degradation from metric points provided in input, where each metric point is annotated with the operator (e.g. phase) it is associated with and the type of metric.

Additionally, tests have been created to ensure the functionality is working as expected. The project structure has also been updated to accommodate these new additions.

The change does not trigger build testing yet.

Fix #90

This change adds a baisc implementation for computing degradataion
metric as described in the paper. The implementation, along with the
helper class, can compute degradation from metric points provided in
input, where each metric point is annotated with the operator (e.g.
phase) it is associated with and the type of metric.

Additionally, tests have been created to ensure the functionality
is working as expected. The project structure has also been updated
to accommodate these new additions.
Comment thread metrics/setup.py
version='0.1',
packages=find_packages(),
install_requires=[
'pandas>=1.5.3',

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I'm not that familiar with Python conventions, but the project contains a list of required packages that were added by Pooja Nilangekar (@poojanilangekar) :
https://github.com/microsoft/lst-bench/blob/main/metrics/notebooks/requirements.txt
Shouldn't this list match that one (and have a single source of truth rather than two as well)?

@@ -0,0 +1,83 @@
"""

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Shouldn't this file be under metrics/utils/ to match the existing folder structure of the Python project? It seems that all the other metrics (cluster, storage) where committed there. If we want to have a folder per metric type, maybe the others should be moved into their own folder (cc Anja Gruenheid (@anjagruenheid) )? WDYT?

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

enhancement New feature or request

Projects

None yet

Development

Successfully merging this pull request may close these issues.

Implement Degradation Metric Calculation for Stability Analysis

2 participants