30
eps : floating point number of the appropriate type
31
The smallest representable number such that ``1.0 + eps != 1.0``.
31
The smallest representable positive number such that
32
``1.0 + eps != 1.0``. Type of `eps` is an appropriate floating
32
34
epsneg : floating point number of the appropriate type
33
The smallest representable number such that ``1.0 - epsneg != 1.0``.
35
The smallest representable positive number such that
36
``1.0 - epsneg != 1.0``.
35
38
The number of bits in the exponent portion of the floating point
38
The object which calculated these parameters and holds more detailed
41
The object which calculated these parameters and holds more
41
The exponent that yields ``eps``.
44
The exponent that yields `eps`.
42
45
max : floating point number of the appropriate type
43
46
The largest representable number.
46
49
min : floating point number of the appropriate type
47
50
The smallest representable number, typically ``-max``.
49
The most negative power of the base (2) consistent with there being
50
no leading 0's in the mantissa.
52
The most negative power of the base (2) consistent with there
53
being no leading 0's in the mantissa.
52
The exponent that yields ``epsneg``.
55
The exponent that yields `epsneg`.
54
57
The number of bits in the exponent including its sign and bias.
56
59
The number of bits in the mantissa.
58
The approximate number of decimal digits to which this kind of float
61
The approximate number of decimal digits to which this kind of
60
63
resolution : floating point number of the appropriate type
61
The approximate decimal resolution of this type, i.e.
64
The approximate decimal resolution of this type, i.e.,
62
65
``10**-precision``.
63
tiny : floating point number of the appropriate type
64
The smallest-magnitude usable number.
67
The smallest positive usable number. Type of `tiny` is an
68
appropriate floating point type.
68
dtype : floating point type, dtype, or instance
69
The kind of floating point data type to get information about.
72
dtype : float, dtype, or instance
73
Kind of floating point data-type about which to get information.
78
For developers of NumPy: do not instantiate this at the module level. The
79
initial calculation of these parameters is expensive and negatively impacts
80
import times. These objects are cached, so calling ``finfo()`` repeatedly
81
inside your functions is not a problem.
82
For developers of NumPy: do not instantiate this at the module level.
83
The initial calculation of these parameters is expensive and negatively
84
impacts import times. These objects are cached, so calling ``finfo()``
85
repeatedly inside your functions is not a problem.