Seems to me that BigDecimal, being essentially a BigInteger numerator and a power-of-ten denominator, could have been accommodated in Category 1 given that Ratio could.
On Mon, Jan 27, 2014 at 10:26 AM, Andy Fingerhut <andy.finger...@gmail.com>wrote: > As Jim already mentioned, you can use == to compare numbers for equality, > but you must be cautious with equality for floating point numbers, as the > tiniest bit of roundoff error will cause = and == to be false for such > comparisons. > > For =, there are effectively 3 'categories' of numeric values in Clojure, > each of which can be = to each other within a category, but between > categories values are never = to each other. > > Category 1. integer values (including Java Byte, Short, Integer, Long, > BigInteger, and Clojure BigInt types) and ratio values (Clojure's Ratio > type) > Category 2. float and double values (Java Float and Double) > Category 3. BigDecimal > > This is better than Java equals(), where Byte and Short are never equals() > to each other, etc. Two numeric values must be the same Java class for > equals() to be true. > > Why the 3 categories, you may wonder? I didn't design it myself, but my > best guess is that in order to have a hash function (in this case Clojure's > hash, implemented as a Java method called hasheq()) that is consistent with > Clojure =, it is difficult to make such a hash function correct and fast if > there were not these 3 separate categories. > > Andy > > > On Mon, Jan 27, 2014 at 5:39 AM, Eric Le Goff <eleg...@gmail.com> wrote: > >> Newbie question : >> >> user=> (= 42 42) >> true >> user=> (= 42 42.0) >> false >> >> I did not expect last result. >> >> I understand that underlying classes are not the same >> i.e >> user=> (class 42) >> java.lang.Long >> user=> (class 42.0) >> java.lang.Double >> >> >> but anyway I'am surprised >> >> Cheers >> >> >> -- >> Eric >> >> -- >> -- >> You received this message because you are subscribed to the Google >> Groups "Clojure" group. >> To post to this group, send email to clojure@googlegroups.com >> Note that posts from new members are moderated - please be patient with >> your first post. >> To unsubscribe from this group, send email to >> clojure+unsubscr...@googlegroups.com >> For more options, visit this group at >> http://groups.google.com/group/clojure?hl=en >> --- >> You received this message because you are subscribed to the Google Groups >> "Clojure" group. >> To unsubscribe from this group and stop receiving emails from it, send an >> email to clojure+unsubscr...@googlegroups.com. >> For more options, visit https://groups.google.com/groups/opt_out. >> > > -- > -- > You received this message because you are subscribed to the Google > Groups "Clojure" group. > To post to this group, send email to clojure@googlegroups.com > Note that posts from new members are moderated - please be patient with > your first post. > To unsubscribe from this group, send email to > clojure+unsubscr...@googlegroups.com > For more options, visit this group at > http://groups.google.com/group/clojure?hl=en > --- > You received this message because you are subscribed to the Google Groups > "Clojure" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to clojure+unsubscr...@googlegroups.com. > For more options, visit https://groups.google.com/groups/opt_out. > -- -- You received this message because you are subscribed to the Google Groups "Clojure" group. To post to this group, send email to clojure@googlegroups.com Note that posts from new members are moderated - please be patient with your first post. To unsubscribe from this group, send email to clojure+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/clojure?hl=en --- You received this message because you are subscribed to the Google Groups "Clojure" group. To unsubscribe from this group and stop receiving emails from it, send an email to clojure+unsubscr...@googlegroups.com. For more options, visit https://groups.google.com/groups/opt_out.
