Re^2: float values and operators

That's what I thought, too, but it doesn't appear to be the case:

$ perl -le '$a = 36.8; $b = 36.6; print "true\n" if $a >= ($b + 0.2);'
# Prints nothing, just as poster showed

$ perl -le '$a = 36.8; $b = 36.6; $c = $b + 0.2; print $c;'
36.8

$ perl -le '$a = 36.8; $b = 36.6; $c = $b + 0.2; print $a;'
36.8

$ perl -le '$a = 36.8; $b = 36.6; $c = $b + 0.2; print $b;'
36.6
[download]

"There is no shame in being self-taught, only in not trying to learn in the first place." -- Atrus, Myst: The Book of D'ni.

Comment on Re^2: float values and operators Download Code

Replies are listed 'Best First'.
Re^3: float values and operators by Zaxo (Archbishop) on Aug 11, 2004 at 20:17 UTC
`$ perl -e'print 36.8 - (36.6 + 0.2)' -7.105427357601e-15` [download] After Compline, Zaxo	[reply] [d/l]
Re^4: float values and operators by hardburn (Abbot) on Aug 11, 2004 at 20:21 UTC
Hrm, that's what I get, too. So why doesn't the imprecision show up when you print the variables out? "There is no shame in being self-taught, only in not trying to learn in the first place." -- Atrus, Myst: The Book of D'ni.	[reply]
Re^5: float values and operators by dave_the_m (Monsignor) on Aug 11, 2004 at 21:11 UTC
Hrm, that's what I get, too. So why doesn't the imprecision show up when you print the variables out? Because print doesn't print the full precision of its args. That's what printf is for: `$ perl585 -e'print 26.6+0.2, "\n"' 26.8 $ perl585 -e'printf "%.20f\n ", 26.6+0.2' 26.80000000000000071054` [download] Dave.	[reply] [d/l]
Re^5: float values and operators by ysth (Canon) on Aug 11, 2004 at 21:56 UTC
Because the default is prudently to print using DBL_DIG or LDBL_DIG digits which should be one less than the number of digits necessary to distinguish any possible value. If you feel exceptionally daring, you can override this by configuring perl with something like -Accflags=-DOVR_DBL_DIG=16 or -Accflags=-DOVR_LDBL_DIG=19. I wouldn't recommend it, though.	[reply]