numpy.finfo() function – Python

numpy.finfo() function shows machine limits for floating point types.

Syntax : numpy.finfo(dtype)

Parameters :
dtype : [float, dtype, or instance] Kind of floating point data-type about which to get information.
Return : Machine parameters for floating point types.

Code #1 :

# Python program explaining
# numpy.finfo() function

# importing numpy as geek
import numpy as geek

gfg = geek.finfo(geek.float32)

print (gfg)

Output :

Machine parameters for float32
---------------------------------------------------------------
precision =   6   resolution = 1.0000000e-06
machep =    -23   eps =        1.1920929e-07
negep =     -24   epsneg =     5.9604645e-08
minexp =   -126   tiny =       1.1754944e-38
maxexp =    128   max =        3.4028235e+38
nexp =        8   min =        -max
---------------------------------------------------------------

Code #2 :

# Python program explaining
# numpy.finfo() function

# importing numpy as geek
import numpy as geek

gfg = geek.finfo(geek.float64)

print (gfg)

Output :

Machine parameters for float64
---------------------------------------------------------------
precision =  15   resolution = 1.0000000000000001e-15
machep =    -52   eps =        2.2204460492503131e-16
negep =     -53   epsneg =     1.1102230246251565e-16
minexp =  -1022   tiny =       2.2250738585072014e-308
maxexp =   1024   max =        1.7976931348623157e+308
nexp =       11   min =        -max
---------------------------------------------------------------