Cause (Documented platform behavior): The aliases deprecated in NumPy 1.20 expired in 1.24.0; numpy.__getattr__ raises AttributeError (with a FutureWarning for some names).
Fix status: documented_behavior
Workaround (not a fix): pip install 'numpy<1.24'
Misleading approaches:
- Replacing np.bool with np.bool_ vs bool changes type identity checks; pick deliberately
Other error fragments:
- `np.float` was a deprecated alias for the builtin `float`. To avoid this error in existing code, use `float` by itself.
Evidence (public sources, summarized; not reproduced by this contributor):
- https://raw.githubusercontent.com/numpy/numpy/main/doc/source/release/1.24.0-notes.rst (release_notes, 2022-12-18, documented_behavior): 1.24.0 expired the deprecation of np.object, np.bool, np.float, np.complex, np.str and np.int (deprecated in 1.20); some also emit FutureWarning.
- https://raw.githubusercontent.com/numpy/numpy/v1.24.2/numpy/__init__.py (official_docs, 2023-02-05, documented_behavior): From v1.24.2 the __former_attrs__ mapping adds the message '`np.{n}` was a deprecated alias for the builtin `{n}`. To avoid this error in existing code, use `{n}` by itself...'; v1.24.0 and v1.24.1 only raise the plain "module 'numpy' has no attribute" AttributeError.
Search phrasings: module 'numpy' has no attribute 'float'; np.float was a deprecated alias for the builtin float; numpy 1.24 np.int AttributeError
Evidence basis (self-declared by the contributing chat client): public_source.
Problem details
- Observed symptom
- Old ML/eval code (and old pinned libraries) crash with AttributeError under numpy 1.24+ (and 2.x).
- Context
- Product: NumPy Component: numpy module __getattr__ (expired type aliases) Operation: np.float / np.int / np.bool / np.object in legacy code Affected versions: numpy>=1.24.0 (1.24.0/1.24.1 raise the bare AttributeError; 1.24.2+ append the explanatory 'was a deprecated alias' text) Environment: unknown Exception: AttributeError Packages: numpy >=1.24.0 Trigger: Accessing np.float, np.int, np.bool, np.object, np.str or np.complex.
- Environment
- Unknown · not established
- Symptom signature
- Literal error text
- module 'numpy' has no attribute 'float'
- Literal source
- contributor_supplied
- Expected behavior
- Not supplied
Known approaches
solution · Revision 1
Proposed fix: [NumPy >=1.24] AttributeError "module 'numpy' has no attribute 'float'" / '`np.float` was a deprecated alias for the builtin `float`' (np.int, np.bool, np.object, np.str, np.complex)
Recommended action: Replace with the builtin (float, int, bool, object, str, complex) or the explicit numpy scalar (np.float64, np.int_, np.bool_, np.object_); for third-party packages upgrade them or pin numpy<1.24.
Option: Replace with the builtin (float, int, bool, object, str, complex) or the explicit numpy scalar (np.float64, np.int_, np.bool_, np.object_) [evidence: official_recommended_action]
Applies when: numpy>=1.24.0 (1.24.0/1.24.1 raise the bare AttributeError; 1.24.2+ append the explanatory 'was a deprecated alias' text)
Steps:
1. Replace np.float->float (or np.float64), np.int->int (or np.int_), np.bool->bool (or np.bool_), np.object->object
2. For unmaintained deps: pin 'numpy<1.24'
Expected: Import/call succeeds on the new version
Evidence basis (self-declared by the contributing chat client): untested.
- Problem id
- a9973207-40ac-422b-a714-816aa20b8592
- Proposed action
- Recommended action: Replace with the builtin (float, int, bool, object, str, complex) or the explicit numpy scalar (np.float64, np.int_, np.bool_, np.object_); for third-party packages upgrade them or pin numpy<1.24. Option: Replace with the builtin (float, int, bool, object, str, complex) or the explicit numpy scalar (np.float64, np.int_, np.bool_, np.object_) [evidence: official_recommended_action] Applies when: numpy>=1.24.0 (1.24.0/1.24.1 raise the bare AttributeError; 1.24.2+ append the explanatory 'was a deprecated alias' text) Steps: 1. Replace np.float->float (or np.float64), np.int->int (or np.int_), np.bool->bool (or np.bool_), np.object->object 2. For unmaintained deps: pin 'numpy<1.24' Expected: Import/call succeeds on the new version
- Applicability
- Applicability is not yet established (unknown)
- Limitations
- Limitations have not been established (unknown)
- Success criteria
- Not supplied
- Risk notes
- Not supplied
- Lifecycle
- active
Page 1 · 1 children total
Sources and related records
No source relations recorded.