 Test of Z ** (1,0) vs Z


   2000 Random arguments were tested from the box
      (     0.1000E+01,     0.1000E+02)
      (     0.0000E+00,     0.1000E+02)


 First, report the vector error


 Test function was larger     0 times,
                   agreed  1347 times, and
              was smaller   653 times.


 There are  53 base   2 significant digits in a floating-point number.


 The maximum relative error of     0.2401E-15 =    2 ** -51.89
    occurred for Z =     0.739900E+01     0.000000E+00
    with a value of      0.739900E+01     0.000000E+00
 The estimated loss of base   2 significant digits is   1.11


 The root mean square relative error was     0.9593E-16 =    2 ** -53.21
 The estimated loss of base   2 significant digits is   0.00


 Then, report the error in the real component.


 Test function was larger   304 times,
                   agreed  1347 times, and
              was smaller   349 times.


 There are  53 base   2 significant digits in a floating-point number.


 The maximum relative error of     0.2401E-15 =    2 ** -51.89
    occurred for Z =     0.739900E+01     0.000000E+00
    with a value of      0.739900E+01     0.000000E+00
 The estimated loss of base   2 significant digits is   1.11


 The root mean square relative error was     0.9593E-16 =    2 ** -53.21
 The estimated loss of base   2 significant digits is   0.00


 Then, report the error in the imaginary component.


 Test function was larger     0 times,
                   agreed     0 times, and
              was smaller  2000 times.


 There are  53 base   2 significant digits in a floating-point number.


 The maximum relative error of     0.1000E+01 =    2 **   0.00
    occurred for Z =     0.100000E+01     0.000000E+00
    with a value of      0.100000E+01     0.000000E+00
 The estimated loss of base   2 significant digits is  53.00


 The root mean square relative error was     0.1000E+01 =    2 **   0.00
 The estimated loss of base   2 significant digits is  53.00


 Test of Z ** (2,0) vs Z*Z


   2000 Random arguments were tested from the box
      (     0.1000E+01,     0.1000E+02)
      (     0.0000E+00,     0.1000E+02)


 First, report the vector error


 Test function was larger     0 times,
                   agreed   801 times, and
              was smaller  1199 times.


 There are  53 base   2 significant digits in a floating-point number.


 The maximum relative error of     0.5164E-15 =    2 ** -50.78
    occurred for Z =     0.908650E+01     0.000000E+00
    with a value of      0.825645E+02     0.000000E+00
 The estimated loss of base   2 significant digits is   2.22


 The root mean square relative error was     0.1713E-15 =    2 ** -52.37
 The estimated loss of base   2 significant digits is   0.63


 Then, report the error in the real component.


 Test function was larger   631 times,
                   agreed   801 times, and
              was smaller   568 times.


 There are  53 base   2 significant digits in a floating-point number.


 The maximum relative error of     0.5164E-15 =    2 ** -50.78
    occurred for Z =     0.908650E+01     0.000000E+00
    with a value of      0.825645E+02     0.000000E+00
 The estimated loss of base   2 significant digits is   2.22


 The root mean square relative error was     0.1713E-15 =    2 ** -52.37
 The estimated loss of base   2 significant digits is   0.63


 Then, report the error in the imaginary component.


 Test function was larger     0 times,
                   agreed     0 times, and
              was smaller  2000 times.


 There are  53 base   2 significant digits in a floating-point number.


 The maximum relative error of     0.1000E+01 =    2 **   0.00
    occurred for Z =     0.100000E+01     0.000000E+00
    with a value of      0.100000E+01     0.000000E+00
 The estimated loss of base   2 significant digits is  53.00


 The root mean square relative error was     0.1000E+01 =    2 **   0.00
 The estimated loss of base   2 significant digits is  53.00


 Test of Z**W vs (Z*Z) ** (W/2)


 *****************************************************
 * Very large MRE values are the norm for this test; *
 * the frequency of exact agreement is meaningful.   *
 *****************************************************


   2000 Random arguments were tested from the box
      (     0.4000E+01,     0.1000E+02)
      (     0.4000E+01,     0.1000E+02)


 First, report the vector error


 Test function was larger     0 times,
                   agreed  1410 times, and
              was smaller   590 times.


 There are  53 base   2 significant digits in a floating-point number.


 The maximum relative error of     0.4082E-14 =    2 ** -47.80
    occurred for Z =     0.756400E+01     0.400000E+01
             and W =     0.400000E+01     0.400000E+01
    with a value of     -0.342102E+03    -0.685229E+03
 The estimated loss of base   2 significant digits is   5.20


 The root mean square relative error was     0.1091E-14 =    2 ** -49.70
 The estimated loss of base   2 significant digits is   3.30


 Then, report the error in the real component.


 Test function was larger   293 times,
                   agreed  1410 times, and
              was smaller   297 times.


 There are  53 base   2 significant digits in a floating-point number.


 The maximum relative error of     0.3414E-11 =    2 ** -38.09
    occurred for Z =     0.983500E+01     0.400000E+01
             and W =     0.400000E+01     0.400000E+01
    with a value of     -0.141015E+01    -0.271034E+04
 The estimated loss of base   2 significant digits is  14.91


 The root mean square relative error was     0.9147E-13 =    2 ** -43.31
 The estimated loss of base   2 significant digits is   9.69


 Then, report the error in the imaginary component.


 Test function was larger   298 times,
                   agreed  1413 times, and
              was smaller   289 times.


 There are  53 base   2 significant digits in a floating-point number.


 The maximum relative error of     0.4714E-14 =    2 ** -47.59
    occurred for Z =     0.637600E+01     0.400000E+01
             and W =     0.400000E+01     0.400000E+01
    with a value of     -0.214740E+03    -0.265271E+03
 The estimated loss of base   2 significant digits is   5.41


 The root mean square relative error was     0.1083E-14 =    2 ** -49.71
 The estimated loss of base   2 significant digits is   3.29


 Special tests


 The identity  Z**W = (1/Z)**(-W)  will be tested.

           Z                     W          (Z**W - (1/Z)**(-W))/CABS(Z**W)

 0.0000E+00 0.0000E+00 0.0000E+00 0.0000E+00??????????????????????????????

 0.0000E+00 0.0000E+00 0.0000E+00 0.0000E+00??????????????????????????????

 0.0000E+00 0.0000E+00 0.0000E+00 0.0000E+00??????????????????????????????

 0.0000E+00 0.0000E+00 0.0000E+00 0.0000E+00??????????????????????????????

 0.0000E+00 0.0000E+00 0.0000E+00 0.0000E+00??????????????????????????????



 Test near an extreme argument.  This should not, but may
 trigger an error message


 (BETA,BETA*XMIN)**(MAXEXP-1,0)       = (  0.8988466+308  0.2046000E+04)
 BETA**(MAXEXP-1)*(1,(MAXEXP-1)*XMIN) = (  0.8988466+308  0.2046000E+04)
 Difference                           = (  0.2814135+295  0.6411938E-10)


 Test of error returns


 Test of (0,0) ** (0,0).
 This should trigger an error message


 Test returned the value (??????????????????????????????)


 This concludes the tests
