ponder-lab / Hybridize-Functions-Refactoring / #1581
72%

DEFAULT BRANCH: main
Ran
Jobs 1
Files 31
Run time 1min
Badge
Embed ▾
README BADGES
x

If you need to use a raster PNG badge, change the '.svg' to '.png' in the link

Markdown

Textile

RDoc

HTML

Rst
coverage: 72.219%. Remained the same
#1581

push

travis-pro

web-flow 
Extend precision audit to fixtures 94-99 (`sparse.eye` and `linalg.eye`) (#534)

* Extend precision audit to fixtures 83-93 (`random.*` family)

Cover the `tf.random.*` constructors with the precision-audit helper. All eleven fixtures are IDEAL—Layer 1 (Ariadne) and Layer 2 (Hybridize inference) both match the Python runtime ground truth, no TODO anchors needed.

- **`random.gamma(shape=[10], alpha=[2 floats])`** (83-86): output `(10, 2)` from `shape + broadcast(alpha)`, FLOAT32 default dtype.
- **`random.normal(shape=[4], mean, stddev, tf.float32)`** (87-89): output `(4,)`, FLOAT32 from explicit dtype argument.
- **`random.poisson(shape=[10], lam=[2 floats])`** (90-91): output `(10, 2)` matching `random.gamma`'s broadcasting rule.
- **`random.truncated_normal(shape=[2], ...)`** (92-93): output `(2,)`, FLOAT32 default dtype.

Add the one-arg `testHasLikelyTensorParameterHelper(TensorType expected)` overload back (it was removed when no caller existed); this batch is the first to use it for the symmetric-Layer-1==Layer-2 case.

Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>

* Extend precision audit to fixtures 94-99 (`sparse.eye` and `linalg.eye`)

Six fixtures across two TF APIs. Numerically all six are IDEAL (Layer 1 and Layer 2 both match the Python runtime ground truth), but the sparse half surfaces a runtime-semantics finding worth anchoring.

**`linalg.eye(2)`** (97-99) — dense identity matrix, `FLOAT32, (2, 2)`. Clean IDEAL case.

**`sparse.eye(2, 3)`** (94-96) — Ariadne emits `FLOAT32, (2, 3)` matching the static shape exactly, and the inference algorithm passes it through unchanged. The numbers are right. The runtime semantics aren't: a `tf.function(input_signature=[TensorSpec(shape=(2, 3), dtype=float32)])` rejects a `SparseTensor` argument, since TF distinguishes `SparseTensor` from `Tensor` at the API level. The precise emission is `tf.SparseTensorSpec`, sibling to the `RaggedTensorSpec` case (#524).

Filed as #533 with `SparseTen... (continued)

1188 of 1645 relevant lines covered (72.22%)

0.72 hits per line

Jobs
ID Job ID Ran Files Coverage
1 #1581.1 31
72.22
Source Files on build #1581