Add tests/parallel/examples/scan/scan_applications.py

[oleksandr-pavlyk]

Description

This file contains two applications.

1. inclusive_segmented_scan_example

This example demonstrates how to compute segmented scan using ordinary scan with crafter scan operation over flag-value struct.

This example uses ZipIterator to efficiently read values and head flags.

2. logcdf_from_logpdf_example

This example uses inclusive_scan with logaddsum operation: $v_1, v_2 \to \log(\exp(v_1) + \exp(v_2))$

Applied to a sequence of log-probabilities, computed for binomial distribution in this example using cupyx, the result should be a strictly increasing sequence that ends with 0.0 == log(1.0).

Due to numerical accuracy issues (?) this is not always true, so we follow up with another application of inclusive_scan with maximum operator, producing strictly non-decreasing sequence.

cp.searchsorted is used to compute quantiles of binomial distribution. Result is compared with reference scipy.stats.distributions.binom.isf

Checklist

• New or existing tests cover these changes.
• The documentation is up to date with these changes.

[github-actions]

🟩 CI finished in 33m 25s: Pass: 100%/22 | Total: 3h 51m | Avg: 10m 31s | Max: 19m 39s

• 🟩 python: Pass: 100%/22 | Total: 3h 51m | Avg: 10m 31s | Max: 19m 39s

  🟩 cpu
    🟩 amd64              Pass: 100%/22  | Total:  3h 51m | Avg: 10m 31s | Max: 19m 39s
  🟩 ctk
    🟩 12.5               Pass: 100%/6   | Total: 46m 36s | Avg:  7m 46s | Max: 15m 34s
    🟩 12.8               Pass: 100%/2   | Total: 39m 01s | Avg: 19m 30s | Max: 19m 38s
    🟩 12.9               Pass: 100%/14  | Total:  2h 26m | Avg: 10m 25s | Max: 19m 39s
  🟩 cudacxx
    🟩 nvcc12.5           Pass: 100%/6   | Total: 46m 36s | Avg:  7m 46s | Max: 15m 34s
    🟩 nvcc12.8           Pass: 100%/2   | Total: 39m 01s | Avg: 19m 30s | Max: 19m 38s
    🟩 nvcc12.9           Pass: 100%/14  | Total:  2h 26m | Avg: 10m 25s | Max: 19m 39s
  🟩 cudacxx_family
    🟩 nvcc               Pass: 100%/22  | Total:  3h 51m | Avg: 10m 31s | Max: 19m 39s
  🟩 cxx
    🟩 GCC13              Pass: 100%/22  | Total:  3h 51m | Avg: 10m 31s | Max: 19m 39s
  🟩 cxx_family
    🟩 GCC                Pass: 100%/22  | Total:  3h 51m | Avg: 10m 31s | Max: 19m 39s
  🟩 gpu
    🟩 h100               Pass: 100%/4   | Total: 41m 21s | Avg: 10m 20s | Max: 17m 49s
    🟩 l4                 Pass: 100%/18  | Total:  3h 10m | Avg: 10m 34s | Max: 19m 39s
  🟩 jobs
    🟩 Build cuda.cccl    Pass: 100%/2   | Total: 19m 59s | Avg:  9m 59s | Max: 10m 10s
    🟩 Test cuda.cccl.cooperative Pass: 100%/5   | Total:  1h 10m | Avg: 14m 01s | Max: 15m 34s
    🟩 Test cuda.cccl.examples Pass: 100%/5   | Total: 24m 13s | Avg:  4m 50s | Max:  5m 56s
    🟩 Test cuda.cccl.headers Pass: 100%/5   | Total: 21m 25s | Avg:  4m 17s | Max:  5m 14s
    🟩 Test cuda.cccl.parallel Pass: 100%/5   | Total:  1h 35m | Avg: 19m 11s | Max: 19m 39s
  🟩 py_version
    🟩 3.10               Pass: 100%/9   | Total:  1h 36m | Avg: 10m 40s | Max: 19m 39s
    🟩 3.13               Pass: 100%/13  | Total:  2h 15m | Avg: 10m 25s | Max: 19m 38s
  

👃 Inspect Changes

Modifications in project?

	Project
	CCCL Infrastructure
	CCCL Packaging
	libcu++
	CUB
	Thrust
	CUDA Experimental
	stdpar
+/-	python
	CCCL C Parallel Library
	Catch2Helper

Modifications in project or dependencies?

	Project
	CCCL Infrastructure
	CCCL Packaging
	libcu++
	CUB
	Thrust
	CUDA Experimental
	stdpar
+/-	python
	CCCL C Parallel Library
	Catch2Helper

🏃‍ Runner counts (total jobs: 22)

#	Runner
16	linux-amd64-gpu-l4-latest-1
4	linux-amd64-gpu-h100-latest-1
2	linux-amd64-cpu16

[leofang]

Very cool!

[shwina]

Very cool indeed!

[github-actions]

🟩 CI finished in 31m 40s: Pass: 100%/22 | Total: 3h 59m | Avg: 10m 52s | Max: 20m 43s

• 🟩 python: Pass: 100%/22 | Total: 3h 59m | Avg: 10m 52s | Max: 20m 43s

  🟩 cpu
    🟩 amd64              Pass: 100%/22  | Total:  3h 59m | Avg: 10m 52s | Max: 20m 43s
  🟩 ctk
    🟩 12.5               Pass: 100%/6   | Total: 46m 18s | Avg:  7m 43s | Max: 13m 42s
    🟩 12.8               Pass: 100%/2   | Total: 40m 45s | Avg: 20m 22s | Max: 20m 35s
    🟩 12.9               Pass: 100%/14  | Total:  2h 32m | Avg: 10m 52s | Max: 20m 43s
  🟩 cudacxx
    🟩 nvcc12.5           Pass: 100%/6   | Total: 46m 18s | Avg:  7m 43s | Max: 13m 42s
    🟩 nvcc12.8           Pass: 100%/2   | Total: 40m 45s | Avg: 20m 22s | Max: 20m 35s
    🟩 nvcc12.9           Pass: 100%/14  | Total:  2h 32m | Avg: 10m 52s | Max: 20m 43s
  🟩 cudacxx_family
    🟩 nvcc               Pass: 100%/22  | Total:  3h 59m | Avg: 10m 52s | Max: 20m 43s
  🟩 cxx
    🟩 GCC13              Pass: 100%/22  | Total:  3h 59m | Avg: 10m 52s | Max: 20m 43s
  🟩 cxx_family
    🟩 GCC                Pass: 100%/22  | Total:  3h 59m | Avg: 10m 52s | Max: 20m 43s
  🟩 gpu
    🟩 h100               Pass: 100%/4   | Total: 45m 10s | Avg: 11m 17s | Max: 17m 00s
    🟩 l4                 Pass: 100%/18  | Total:  3h 14m | Avg: 10m 47s | Max: 20m 43s
  🟩 jobs
    🟩 Build cuda.cccl    Pass: 100%/2   | Total: 19m 18s | Avg:  9m 39s | Max:  9m 40s
    🟩 Test cuda.cccl.cooperative Pass: 100%/5   | Total:  1h 11m | Avg: 14m 18s | Max: 16m 25s
    🟩 Test cuda.cccl.examples Pass: 100%/5   | Total: 26m 11s | Avg:  5m 14s | Max:  6m 18s
    🟩 Test cuda.cccl.headers Pass: 100%/5   | Total: 23m 14s | Avg:  4m 38s | Max:  5m 27s
    🟩 Test cuda.cccl.parallel Pass: 100%/5   | Total:  1h 39m | Avg: 19m 49s | Max: 20m 43s
  🟩 py_version
    🟩 3.10               Pass: 100%/9   | Total:  1h 37m | Avg: 10m 48s | Max: 20m 39s
    🟩 3.13               Pass: 100%/13  | Total:  2h 22m | Avg: 10m 56s | Max: 20m 43s
  

👃 Inspect Changes

Modifications in project?

	Project
	CCCL Infrastructure
	CCCL Packaging
	libcu++
	CUB
	Thrust
	CUDA Experimental
	stdpar
+/-	python
	CCCL C Parallel Library
	Catch2Helper

Modifications in project or dependencies?

	Project
	CCCL Infrastructure
	CCCL Packaging
	libcu++
	CUB
	Thrust
	CUDA Experimental
	stdpar
+/-	python
	CCCL C Parallel Library
	Catch2Helper

🏃‍ Runner counts (total jobs: 22)

#	Runner
16	linux-amd64-gpu-l4-latest-1
4	linux-amd64-gpu-h100-latest-1
2	linux-amd64-cpu16