Raffi Khatchadourian

Make `com.ibm.wala.cast.python.ml.test` use Jython3

2

This also makes me believe that the ML tests are using Jython and not Jython3. _Originally posted by @khatchad in https://github.com/wala/ML/issues/146#issuecomment-1947657641_

Missing decorated functions whose decorators simply returns the function

4

## Description Related to https://github.com/wala/ML/issues/189. Consider the following code: ```python import tensorflow as tf def mama(fun): return fun @mama def raffi(x): assert isinstance(x, tf.Tensor) raffi(tf.constant(1)) ``` The function `raffi()` is...

Use Java 17.

- **Add tests.** - **Fix comments.** - **Extract method.** - **Capitalize constant.** - **Shorten name.** - **Remove explicit super call.** It's called implicitly.

No argument expansion for custom decorators with function parameters

1

Related to https://github.com/wala/ML/issues/91. Consider the following code: ```python import tensorflow as tf def mama(fun): def wrapper_fun(*args, **kwargs): assert isinstance(args[0], tf.Tensor) fun(*args, **kwargs) return wrapper_fun @mama def f(x): assert isinstance(x, tf.Tensor)...

Track tensors emanating from `tf.data.Dataset.zip()`

1

[`zip()`](https://www.tensorflow.org/versions/r2.9/api_docs/python/tf/data/Dataset#zip) looks very similar. _Originally posted by @khatchad in https://github.com/wala/ML/issues/165#issuecomment-1989296987_

Track individual tensors in containers (datasets, tuples, etc)

2

Currently, we track that datasets have tensors, but we don't track the tensors used to make the datasets. For example, the `from_tensors()` static method creates a dataset containing a single...

Track tensors emanating from `tf.data.Dataset.enumerate()`

2

Since https://github.com/wala/ML/pull/132, we can `enumerate()` over datasets and track the tensors. However, the TF Data API has a method [`tf.data.Dataset.enumerate()`](https://www.tensorflow.org/versions/r2.9/api_docs/python/tf/data/Dataset#enumerate) that does is similar to Python's `enumerate()`. But, we don't...

Consider the following code: ```python from tensorflow import Tensor def f(a): assert isinstance(a, Tensor) ``` I am seeing this in the tensor analysis: ``` [Node: synthetic < PythonLoader, Lwala/builtin/isinstance, do()LRoot;...

Consider the two input files below: 1. https://github.com/wala/ML/blob/404a34f5034a2647c512ae96a021a153f7abccd6/com.ibm.wala.cast.python.test/data/multi1.py 2. https://github.com/wala/ML/blob/404a34f5034a2647c512ae96a021a153f7abccd6/com.ibm.wala.cast.python.test/data/multi2.py Note that 1 uses 2. Thus, you must pass 2 to the engine before passing 1. In other words, you...