ajayshah@alhena.usc.edu (Ajay Shah) (05/23/91)
This is the fortran program:
double precision x
integer i
double precision anorm
do i = -30, 20
x = 1.0d0*i
write(6, '(E30.22)') anorm(x)
end do
stop
end
Here are the two outputs. In this demo, nothing serious is
affected but it represents a difference between a fortran program
and a f2c translation. This makes using f2c in more complex
situations difficult.
from f2c from f77
.4906713927148187177321-197 0.4906713927148187177321-197
.3289785266704380867514-184 0.3289785266704380867514-184
.8123869469659427264895-172 0.8123869469659427264895-172
.7389481006885018453671-160 0.7389481006885018453671-160
.2476063315503389241044-148 0.2476063315503389241044-148
.3056696706382561616593-137 0.3056696706382561616593-137
.1390392118549702977187-126 0.1390392118549702977187-126
.2330637006220649159568-116 0.2330637006220649159568-116
.1439892435145079305123-106 0.1439892435145079305123-106
.3279278018979036074213E-97 0.3279278018979036074213E-97
.2753624118606233737134E-88 0.2753624118606233737134E-88
.8527223952630977222754E-80 0.8527223952630977222754E-80
.9740948918937150769086E-72 0.9740948918937150769086E-72
.4105996202098906529308E-64 0.4105996202098906529308E-64
.6388754400538088218520E-57 0.6388754400538088218520E-57
.3670966199312751422570E-50 0.3670966199312751422570E-50
.7793536819192800037878E-44 0.7793536819192800037878E-44
.6117164399549880257194E-38 0.6117164399549880257194E-38
.1776482112077679037849E-32 0.1776482112077679037849E-32
.1910659574498675226475E-27 0.1910659574498675226475E-27
.7619853024160526919908E-23 0.7619853024160526919908E-23
.1128588405953840782741E-18 0.1128588405953840782741E-18
.6220960574271784860433E-15 0.6220960574271784860433E-15
.1279812543885835163949E-11 0.1279812543885835163949E-11
.9865876450376980905188E-09 0.9865876450376980905188E-09
.2866515718791939118538E-06 0.2866515718791939118538E-06
.3167124183311992432696E-04 0.3167124183311992432696E-04
.1349898031630094577191E-02 0.1349898031630094577191E-02
.2275013194817921205471E-01 0.2275013194817921205471E-01
.1586552539314570464679E+00 0.1586552539314570464679E+00
.5000000000000000000000E+00 0.5000000000000000000000E+00
.8413447460685429257765E+00 0.8413447460685429257765E+00
.9772498680518207914147E+00 0.9772498680518207914147E+00
.9986501019683698965324E+00 0.9986501019683698965324E+00
.9999683287581668800215E+00 0.9999683287581668800215E+00
.9999997133484280764648E+00 0.9999997133484280764648E+00
.9999999990134122995755E+00 0.9999999990134122995755E+00
.9999999999987201348972E+00 0.9999999999987201348972E+00
.9999999999999993338662E+00 0.9999999999999993338662E+00
.1000000000000000000000E+01 0.1000000000000000000000E+01
.1000000000000000000000E+01 0.1000000000000000000000E+01
.1000000000000000000000E+01 0.1000000000000000000000E+01
.1000000000000000000000E+01 0.1000000000000000000000E+01
.1000000000000000000000E+01 0.1000000000000000000000E+01
.1000000000000000000000E+01 0.1000000000000000000000E+01
.1000000000000000000000E+01 0.1000000000000000000000E+01
.1000000000000000000000E+01 0.1000000000000000000000E+01
.1000000000000000000000E+01 0.1000000000000000000000E+01
.1000000000000000000000E+01 0.1000000000000000000000E+01
.1000000000000000000000E+01 0.1000000000000000000000E+01
.1000000000000000000000E+01 0.1000000000000000000000E+01
--
_______________________________________________________________________________
Ajay Shah, (213)734-3930, ajayshah@usc.edu
The more things change, the more they stay insane.
_______________________________________________________________________________ags@seaman.cc.purdue.edu (Dave Seaman) (05/23/91)
In article <33089@usc> ajayshah@alhena.usc.edu (Ajay Shah) writes: >Here are the two outputs. In this demo, nothing serious is >affected but it represents a difference between a fortran program >and a f2c translation. This makes using f2c in more complex >situations difficult. > > from f2c from f77 > > .4906713927148187177321-197 0.4906713927148187177321-197 [ many lines deleted ... ] Since you didn't say what difference you thought you saw, it is a little difficult to answer the question. The only obvious difference is that f77 provides the optional zero before the decimal point, and f2c omits it. Both are standard-conforming. -- Dave Seaman ags@seaman.cc.purdue.edu