Raffi Khatchadourian

Actually, I'm unsure whether these docs are for the correct API. I don't see a call to it anywhere in the examples.

Ah, they haven't been updated for TF2. They still refer to the old [`matrix_set_diag`](https://www.tensorflow.org/api_docs/python/tf/linalg/set_diag) API.

> @khatchad Could you please let us know if you have referred to the above [comment](https://github.com/tensorflow/tensorflow/issues/67255#issuecomment-2104241615) and you could see the return information ? Thank you! Not sure if I...

Great, thank you!

This looks like it was part of the original code: https://github.com/wala/ML/blame/261455fb03e274947967022d822caec089c4f5e4/com.ibm.wala.cast.python/source/com/ibm/wala/cast/python/ir/PythonCAstToIRTranslator.java#L615-L617

Related to https://github.com/wala/ML/issues/147.

Add this test works: https://github.com/ponder-lab/ML/commit/2db36e27afd9399c70a788a6800b675c19505379 I believe the problem is that the data sources are hard-coded: https://github.com/wala/ML/blob/ddba21e7881f2a9cc825f1857aea5a5ea89f1bc3/com.ibm.wala.cast.python.ml/source/com/ibm/wala/cast/python/ml/client/PythonTensorAnalysisEngine.java#L606-L610

But, it could also have something to do with the way other APIs are being constructed. For example, the points-to set for `tf.ones()` is empty.

Looking at the summary of `tf.reshape()`, I see that there's a data copy: https://github.com/wala/ML/blob/ddba21e7881f2a9cc825f1857aea5a5ea89f1bc3/com.ibm.wala.cast.python.ml/data/tensorflow.xml#L379-L388 That may mean if the data source has something wrong with it, any copied data would...

That being said, `copy_data()` in the above summary doesn't use its argument.