add correctness data and class size unit test metrics - facilitates certainty in correct data flow

microsoft / seismic-deeplearning

Deep Learning for Seismic Imaging and Interpretation

MIT License

403 stars 142 forks source link

add correctness data and class size unit test metrics - facilitates certainty in correct data flow #317

Closed maxkazmsft closed 4 years ago

maxkazmsft commented 4 years ago

add performance metrics similar to test/cicd/src/check_performance.py to check whether the input data and predicted results on synthetic checkerboard dataset have the correct dimensions, etc - this essentially checks the correctness of data flow through the project

Checking data sizes only - not distributions

number of classes - training, prediction
length of dataset
length of data loader

Pre-requisit #346

fazamani commented 4 years ago

add a test to check the bounds (limits) on the data