Would some good soul please explain what
Number readFrom: '7r6.5' should return. It is apparently not the same as 7r65 * 0.1 and before I dig deeper (why does this happen, is it a bug or feature) I'd appreciate that someone confirms or rejects (bug | feature). Thank you. /Klaus |
On Mon, 2006-07-03 at 20:30 +0200, Klaus D. Witzel wrote:
> Would some good soul please explain what > > Number readFrom: '7r6.5' > > should return. It is apparently not the same as 7r65 * 0.1 and before I > dig deeper (why does this happen, is it a bug or feature) I'd appreciate > that someone confirms or rejects (bug | feature). > > Thank you. > > /Klaus 7r65 * 0.1 the first value is in base 7 whereas the second is in base 10 7r65 * 7r0.1 IS the same valus as 7r6.5. Ken signature.asc (196 bytes) Download Attachment |
Hi Ken,
on Mon, 03 Jul 2006 20:51:00 +0200, you wrote: > On Mon, 2006-07-03 at 20:30 +0200, Klaus D. Witzel wrote: >> Would some good soul please explain what >> >> Number readFrom: '7r6.5' >> >> should return. It is apparently not the same as 7r65 * 0.1 and before I >> dig deeper (why does this happen, is it a bug or feature) I'd appreciate >> that someone confirms or rejects (bug | feature). >> >> Thank you. >> >> /Klaus > > Because in the expression > > 7r65 * 0.1 > > the first value is in base 7 whereas the second is in base 10 > > 7r65 * 7r0.1 IS the same valus as 7r6.5. Ah! This equation cut my knot, thank you Ken. /Klaus > Ken |
In reply to this post by Klaus D. Witzel
Klaus, you asked:
>Would some good soul please explain what > > Number readFrom: '7r6.5' And here is the good soul's answer: 7r6.5 = 6*7^0 + 5 * 7^(-1) = 6 + 5/7 = 6.714285714285... with the last 6 digits repeating at infinitum. Whereas 7r65 * 0.1 should evaluate to 7r65 * 0.1 = (6*7^1 + 5*7^0) * 10^-1 = (6*7 + 5) / 10 = 4.7 (Notation: "^" is the exponentation operator, like 3^2 = 3 * 3 = 9). Greetings, Wolfgang -- "Dijkstra is right, but you don't say such things!" (A less courageous programmer) |
Thank you Wolfgang (did you see the response from Ken).
But one thing remains: in 7r6.5e4 the exponent is also computed base 7, contrary to the comment in NumberParsingTest>>#testFloatReadWithRadix. /Klaus On Tue, 04 Jul 2006 20:11:04 +0200, Wolfgang Helbig wrote: > Klaus, you asked: >> Would some good soul please explain what >> >> Number readFrom: '7r6.5' > > And here is the good soul's answer: > > 7r6.5 = 6*7^0 + 5 * 7^(-1) = 6 + 5/7 = 6.714285714285... with the last > 6 digits repeating at infinitum. > > Whereas 7r65 * 0.1 should evaluate to > 7r65 * 0.1 = (6*7^1 + 5*7^0) * 10^-1 = (6*7 + 5) / 10 = 4.7 > > (Notation: "^" is the exponentation operator, like 3^2 = 3 * 3 = 9). > > Greetings, > Wolfgang > > -- > "Dijkstra is right, but you don't say such things!" > (A less courageous programmer) > > > |
In reply to this post by Klaus D. Witzel
Yes, 7r6.5e4 is equivalent to 7r65000.0 That is the natural behavior we would expect i think To be more precise, however, the number on the right of exponent e is interpreted in base 10. let us take another example: 7r6.5e12 is (6*7^0+5*7^-1) * 7^12 that is 6*7^12+5*7^11 it is not interpreted as (6*7^0+5*7^-1) * 7^ 7r12 that is 6*7^9+5*7^8 Nicolas Klaus D. Witzel: > Thank you Wolfgang (did you see the response from Ken). > > But one thing remains: in 7r6.5e4 the exponent is also computed base 7, > contrary to the comment in NumberParsingTest>>#testFloatReadWithRadix. > > /Klaus > > On Tue, 04 Jul 2006 20:11:04 +0200, Wolfgang Helbig wrote: > > > Klaus, you asked: > >> Would some good soul please explain what > >> > >> Number readFrom: '7r6.5' > > > > And here is the good soul's answer: > > > > 7r6.5 = 6*7^0 + 5 * 7^(-1) = 6 + 5/7 = 6.714285714285... with the last > > 6 digits repeating at infinitum. > > > > Whereas 7r65 * 0.1 should evaluate to > > 7r65 * 0.1 = (6*7^1 + 5*7^0) * 10^-1 = (6*7 + 5) / 10 = 4.7 > > > > (Notation: "^" is the exponentation operator, like 3^2 = 3 * 3 = 9). > > > > Greetings, > > Wolfgang > > > > -- > > "Dijkstra is right, but you don't say such things!" > > (A less courageous programmer) > > > > > > > > > > ________________________________________________________________________ iFRANCE, exprimez-vous ! http://web.ifrance.com |
Free forum by Nabble | Edit this page |