search

tallamjr / astronet

Efficient Deep Learning for Real-time Classification of Astronomical Transients and Multivariate Time-series
Apache License 2.0
14 stars 3 forks source link

Add ability to run analysis on the PLAsTiCC data #25

Closed tallamjr closed 4 years ago

tallamjr commented 4 years ago

This code allows for one to run the astronet.t2 analysis on the PLAsTiCC dataset.

Code for GP interpolation has been added as well as helper functions to transform the dataset into a form akin to wisdm_2010

Closes #20

tallamjr commented 4 years ago

All tests pass locally, but fail due to private repo time limit on Github actions.

375.14s call     astronet/t2/tests/unit/test_utils.py::test_load_plasticc
115.85s call     astronet/t2/tests/int/test_train.py::test_training_pipeline_wisdm_2010
31.47s call     astronet/t2/tests/unit/test_utils.py::test_load_wisdm_2019
29.85s call     astronet/t2/tests/unit/test_preprocess.py::test_one_hot_encode_local
10.13s call     astronet/t2/tests/unit/test_utils.py::test_load_wisdm_2010
10.07s call     astronet/t2/tests/unit/test_preprocess.py::test_one_hot_encode
1.54s call     astronet/t2/tests/unit/test_preprocess.py::test_plasticc_predict_2d_gp
1.51s call     astronet/t2/tests/func/test_gp_interpolation.py::test_plasticc_gp_interpolation
0.02s call     astronet/t2/tests/unit/test_tf_multihead_attention.py::MultiheadAttentionTest::testMultiHeadAttentionOutputShapes
0.01s call     astronet/t2/tests/unit/test_tf_multihead_attention.py::MultiheadAttentionTest::testScaledDotProductAttentionMask
0.01s call     astronet/t2/tests/unit/test_tf_dense_layer.py::DenseLayerTest::testDenseLayerOutput

(40 durations < 0.005s hidden.  Use -vv to show these durations.)
=========================================================================================== 15 passed, 2 skipped, 2 warnings in 581.08s (0:09:41) ============================================================================================