fix(coordinate_compression): raise exceptions instead of returning -1

[AshrafHosam]

Describe your change:

Fixes inconsistent error handling in data_compression/coordinate_compression.py reported in #13509.

Previously:

• compress(x) returned -1 on missing input (silent failure)
• decompress(n) returned -1 on out-of-range input (silent failure)
• coordinate_map[x] raised KeyError (exception)

This mix of sentinels and exceptions made errors easy to miss and hard to debug.

This PR unifies the behavior to raise exceptions consistently:

• compress(x) now raises KeyError (same as the underlying coordinate_map dict)
• decompress(n) now raises IndexError with a descriptive message
• Docstrings updated to document the new Raises: contract
• Doctests updated to cover both the happy path and the error path

Fixes #13509

• Add an algorithm?
• Fix a bug or typo in an existing algorithm?
• Add or change doctests? -- Note: Please avoid changing solely the doctests of an existing algorithm.
• Documentation change?

Checklist:

• I have read CONTRIBUTING.md.
• This pull request is all my own work -- I have not plagiarized.
• I know that pull requests will not be merged if they fail the automated tests.
• This PR only changes one algorithm file. To ease review, please open separate PRs for separate algorithms.
• All new Python files are placed inside an existing directory.
• All filenames are in all lowercase characters with no spaces or dashes.
• All functions and variable names follow Python naming conventions.
• All function parameters and return values are annotated with Python type hints.
• All functions have doctests that pass the automated testing.
• All new algorithms include at least one URL that points to Wikipedia or another similar explanation.
• If this pull request resolves one or more open issues then the description above includes the issue number(s) with a closing keyword: "Fixes #ISSUE-NUMBER".

[cclauss]

@priya-sundaram-dev, your review, please, as a fix to

• Bugs #13509

[priya-sundaram-dev]

Thanks @AshrafHosam — the direction here is right: replacing the silent -1 returns with real exceptions is what #13509 asks for, and the doctests are self-verifying and pass locally (18 tests). Two things to consider before this fully closes the issue:

1. Silent -1 still leaks for duplicate inputs. compress_coordinates() sizes reverse_map as [-1] * len(arr) but only fills one slot per unique value, so duplicates leave trailing -1 sentinels that are still in range for decompress:

>>> cc = CoordinateCompressor([10, 10, 52])
>>> cc.reverse_map
[10, 52, -1]
>>> cc.decompress(2)   # in range (0 <= 2 < 3), so no IndexError
-1

So the "silent failure" this PR removes from the no-duplicates path is reintroduced on the duplicate path. Building the maps from sorted(set(arr)) (so reverse_map has exactly the unique count) closes that hole.

2. Consistency of exception type. The issue title is about a mix of behaviors; compress now raises KeyError while decompress raises IndexError. Both are exceptions (good), but a single type — e.g. ValueError for both — reads as more "consistent" and is easier to document/catch.

Heads-up that #14873 takes the sorted(set(...)) approach and raises ValueError uniformly, which covers both points above — worth a look to align with, or borrow from. Nice work either way.

[priya-sundaram-dev]

LGTM — this is exactly what #13509 asked for. Replacing the -1 sentinels with real exceptions is the right call:

• compress now surfaces a KeyError for unknown values instead of a magic number a caller could silently treat as a valid coordinate.
• decompress raises IndexError with a helpful message rather than returning -1 (which collided with a legitimate value/type domain).

The Raises: sections and the Traceback (most recent call last): doctests are correct — I ran them locally and both the KeyError: 7 and IndexError: compressed coordinate 5 is out of range cases reproduce cleanly. No behavioural gaps: every previous -1 return path is now covered by an exception. Nice, minimal diff.