Add complex number support to cosh

[kgryte]

This PR

• adds complex number support to cosh by documenting special cases. The hyperbolic cosine is an entire function in the complex plane. Thus, the function has no branch cuts.
• updates the input and output array data types to be any floating-point data type, not just real-valued floating-point data types.
• derives special cases from C99 and tested against NumPy (script found below).

import numpy as np
import math

def is_equal_float(x, y):
    """Test whether two floating-point numbers are equal with special consideration for zeros and NaNs.

    Parameters
    ----------
    x : float
        First input number.
    y : float
        Second input number.

    Returns
    -------
    bool
        Boolean indicating whether two floating-point numbers are equal.

    Examples
    --------
    >>> is_equal_float(0.0, -0.0)
    False
    >>> is_equal_float(-0.0, -0.0)
    True
    """
    # Handle +-0:
    if x == 0.0 and y == 0.0:
        return math.copysign(1.0, x) == math.copysign(1.0, y)

    # Handle NaNs:
    if x != x:
        return y != y

    # Everything else, including infinities:
    return x == y


def is_equal(x, y):
    """Test whether two complex numbers are equal with special consideration for zeros and NaNs.

    Parameters
    ----------
    x : complex
        First input number.
    y : complex
        Second input number.

    Returns
    -------
    bool
        Boolean indicating whether two complex numbers are equal.

    Examples
    --------
    >>> import numpy as np
    >>> is_equal(complex(np.nan, np.nan), complex(np.nan, np.nan))
    True
    """
    return is_equal_float(x.real, y.real) and is_equal_float(x.imag, y.imag)


# Strided array consisting of input values and expected values:
values = [
    complex(0.0, 0.0),        # 0
    complex(1.0, 0.0),        # 0
    
    complex(-0.0, 0.0),       # 1
    complex(1.0, -0.0),       # 1
    
    complex(0.0, np.inf),     # 2
    complex(np.nan, 0),       # 2
    
    complex(0.0, np.nan),     # 3
    complex(np.nan, 0.0),     # 3
    
    complex(1.0, np.inf),     # 4
    complex(np.nan, np.nan),  # 4
    
    complex(1.0, np.nan),     # 5
    complex(np.nan, np.nan),  # 5
    
    complex(np.inf, 0.0),     # 6
    complex(np.inf, 0.0),     # 6
    
    complex(np.inf, -0.0),    # 7
    complex(np.inf, -0.0),    # 7
    
    complex(np.inf, 1.0),     # 8
    complex(np.inf, np.inf),  # 8
    
    complex(np.inf, np.inf),  # 9
    complex(np.inf, np.nan),  # 9
    
    complex(np.inf, np.nan),  # 10
    complex(np.inf, np.nan),  # 10
    
    complex(np.nan, 0.0),     # 11
    complex(np.nan, 0.0),     # 11
    
    complex(np.nan, 1.0),     # 12
    complex(np.nan, np.nan),  # 12

    complex(np.nan, 1.0),     # 13
    complex(np.nan, np.nan),  # 13
    
    complex(np.nan, np.nan),  # 14
    complex(np.nan, np.nan)   # 14
]

for i in range(len(values)//2):
    j = i * 2
    v = values[j]
    e = values[j+1]
    actual = np.cosh(v)
    print('Value: {value}'.format(value=str(v)))
    print('Actual: {actual}'.format(actual=str(actual)))
    print('Expected: {expected}'.format(expected=str(e)))
    print('Equal: {is_equal}'.format(is_equal=str(is_equal(actual, e))))
    print('\n')

Value: 0j
Actual: (1+0j)
Expected: (1+0j)
Equal: True


Value: (-0+0j)
Actual: (1-0j)
Expected: (1-0j)
Equal: True


/path/to/ccosh.py:116: RuntimeWarning: invalid value encountered in cosh
  actual = np.cosh(v)
Value: infj
Actual: (nan+0j)
Expected: (nan+0j)
Equal: True


Value: nanj
Actual: (nan+0j)
Expected: (nan+0j)
Equal: True


Value: (1+infj)
Actual: (nan+nanj)
Expected: (nan+nanj)
Equal: True


Value: (1+nanj)
Actual: (nan+nanj)
Expected: (nan+nanj)
Equal: True


Value: (inf+0j)
Actual: (inf+0j)
Expected: (inf+0j)
Equal: True


Value: (inf-0j)
Actual: (inf-0j)
Expected: (inf-0j)
Equal: True


Value: (inf+1j)
Actual: (inf+infj)
Expected: (inf+infj)
Equal: True


Value: (inf+infj)
Actual: (inf+nanj)
Expected: (inf+nanj)
Equal: True


Value: (inf+nanj)
Actual: (inf+nanj)
Expected: (inf+nanj)
Equal: True


Value: (nan+0j)
Actual: (nan+0j)
Expected: (nan+0j)
Equal: True


Value: (nan+1j)
Actual: (nan+nanj)
Expected: (nan+nanj)
Equal: True


Value: (nan+1j)
Actual: (nan+nanj)
Expected: (nan+nanj)
Equal: True


Value: (nan+nanj)
Actual: (nan+nanj)
Expected: (nan+nanj)
Equal: True

[kgryte]

As cosh does not involve branch cuts and the special cases introduced in this PR follow prior art, will merge. Further revisions can be addressed in subsequent PRs...