archimedes.array¶

archimedes.array(x: Any, dtype: DTypeLike | None = None) → ArrayLike¶

Create an array supporting both numeric and symbolic computation.

This function serves as Archimedes’ array creation mechanism, handling both numeric and symbolic inputs. It creates the appropriate array type (NumPy or SymbolicArray) based on the input data.

Parameters:

x (array_like) –
An array-like object, which can be:
- NumPy ndarray
- SymbolicArray
- List or nested list of scalars
- Scalar value
dtype (str or numpy.dtype, optional) – The data type for the array. If not specified, dtype is inferred from x.

Returns:

array – If the input contains symbolic elements, returns a SymbolicArray. Otherwise, returns a NumPy ndarray.

Return type:

numpy.ndarray or SymbolicArray

Notes

Array creation using the NumPy dispatch mechanism (np.array(..., like=...)) is recommended over calling array(...) directly. The dispatch mechanism supports a wider range of input types and better handles numeric input types.

When working with symbolic computation, this function ensures that array creation follows the same patterns as NumPy, allowing for seamless transitions between symbolic and numeric computation.

This function currently supports:

Creating arrays from existing arrays (preserving type)
Creating arrays from lists of scalars (1D arrays)
Creating arrays from lists of lists (2D arrays)

Limitations and edge cases:

Higher-dimensional arrays (>2D) are not supported
Creating arrays with inconsistent dimensions will raise a ValueError
Complex-valued arrays may have limited symbolic operation support

Examples

>>> import numpy as np
>>> import archimedes as arc
>>>
>>> # Basic numeric array creation
>>> arc.array([1.0, 2.0, 3.0])
array([1., 2., 3.])
>>>
>>> # Creating a symbolic array
>>> x = arc.sym("x", 3)  # Create a symbolic variable
>>> arc.array([x[0], 2.0, x[2]])  # Mixed symbolic/numeric content
[x_0, 2, x_2]
>>>
>>> # 2D array from nested lists
>>> arc.array([[1.0, 2.0], [3.0, 4.0]])
array([[1., 2.],
       [3., 4.]])
>>>
>>> # Using NumPy dispatch (recommended approach)
>>> np.array([x[0], 2.0, x[2]], like=x_sym)  # Creates a SymbolicArray
[x_0, 2, x_2]