2 Months..... umm, weeks to go (2012-02-06)


Hmm... I think it's 2.. EVEN my other teacher long ago said .5 is not roundable... :p

hum, i've been teached for 9years that 1,5 was rounded toward 2. that matter was laid to rest for 2years and I was then teached for the 5 following years that 1,5 was rounded to 1...


that said I love your avatar...
 
There's only one reason to round *.5's to the next number all the time: if there are any more digits after the 5 (e.g. 1.5001), it's closer to the higher number. But if the number is exactly *.5, rounding up all the time is simply an arbitrary rule. A necessary one for consistency, but arbitrary nonetheless, so saying 1.5 rounded is 1 is just as valid as saying 1.5 rounded is 2. It's the exact same difference, after all.
 
There's only one reason to round *.5's to the next number all the time: if there are any more digits after the 5 (e.g. 1.5001), it's closer to the higher number. But if the number is exactly *.5, rounding up all the time is simply an arbitrary rule. A necessary one for consistency, but arbitrary nonetheless, so saying 1.5 rounded is 1 is just as valid as saying 1.5 rounded is 2. It's the exact same difference, after all.

computer radically changed that vision...1,9999 generally round to 1 when considering integers...
 
Ok, this conversation just turned waaaaaay to nerdy for me (and I'm a physicist). :p
 
computer radically changed that vision...1,9999 generally round to 1 when considering integers...
That's truncating, not rounding. :p

There's only one reason to round *.5's to the next number all the time: if there are any more digits after the 5 (e.g. 1.5001), it's closer to the higher number. But if the number is exactly *.5, rounding up all the time is simply an arbitrary rule. A necessary one for consistency, but arbitrary nonetheless, so saying 1.5 rounded is 1 is just as valid as saying 1.5 rounded is 2. It's the exact same difference, after all.
Of course it's arbitrary. It's also fairly consistent, with the majority of the world teaching this method (for many reasons) so that when someone describes rounding a perfect midpoint it is supposed to be understood what is meant without having to say it. That's why we have standards, so we can communicate ideas without having to explain exactly what is meant every single time.
 
Of course it's arbitrary. It's also fairly consistent, with the majority of the world teaching this method (for many reasons) so that when someone describes rounding a perfect midpoint it is supposed to be understood what is meant without having to say it. That's why we have standards, so we can communicate ideas without having to explain exactly what is meant every single time.

And for the times we can't be bothered to explain exactly, we have Wikipedia: http://en.wikipedia.org/wiki/Rounding#Tie-breaking :)
 
Ok, this conversation just turned waaaaaay to nerdy for me (and I'm a physicist). :p

you sir are lieing... :D
Ok, fine! You got me <_< ... I'm still following the conversation with high interest :lol:


(And the answer is 1.5+/-0.5, which means that there is no clear detection on the 3 sigma level anyway :p )
 
Last edited by a moderator:
1 and 2 are exactly the same difference from 1.5.
And the rule for rounding .5 that we all should have learned in the first grade is...?

Heh. Actually, here in Sweden, they had some sort of campaign to replace the "always round up" rule with the "Always round to the nearest even number". The reason being that they sort of wanted to redress the inequality of different rounding errors over time. Or something like that. This was some time after they had gotten the idea that kids couldn't understand numbers, and promptly proceeded to converting first grade mathematics into...Set theory!


...oh, how I love political intervention into pedagogy :)
 
There's only one reason to round *.5's to the next number all the time: if there are any more digits after the 5 (e.g. 1.5001), it's closer to the higher number. But if the number is exactly *.5, rounding up all the time is simply an arbitrary rule. A necessary one for consistency, but arbitrary nonetheless, so saying 1.5 rounded is 1 is just as valid as saying 1.5 rounded is 2. It's the exact same difference, after all.
Of course it's arbitrary. It's also fairly consistent, with the majority of the world teaching this method (for many reasons) so that when someone describes rounding a perfect midpoint it is supposed to be understood what is meant without having to say it. That's why we have standards, so we can communicate ideas without having to explain exactly what is meant every single time.
Like jkt mentions, there are other methods of rounding. I conjecture that we generally use the "round up" method (well, actually, the "round away from 0" method) because it means it is possible to accurately round any floating-point number with more digits than the tenths place while only looking at the tenths place; 0.50001 is technically closer to 1 than 0, so rounding up is more accurate than rounding down. On the other hand, if we used the "round to even" rule, we would have to always remember to look past the tenths place and always round up if it isn't exactly *.5, because if we rounded 2.52 to 2, that would be less accurate than rounding it to 3. So rounding up always is a simplification, and a good one. But still, when rounding 1.5, both 1 and 2 are just as accurate choices, mathematically; it's the exact same difference.
 
Last edited by a moderator:
Like jkt mentions, there are other methods of rounding. I conjecture that we generally use the "round up" method (well, actually, the "round away from 0" method) because it means it is possible to accurately round any floating-point number with more digits than the tenths place while only looking at the tenths place; 0.50001 is technically closer to 1 than 0, so rounding up is more accurate than rounding down. On the other hand, if we used the "round to even" rule, we would have to always remember to look past the tenths place and always round up if it isn't exactly *.5, because if we rounded 2.52 to 2, that would be less accurate than rounding it to 3. So rounding up always is a simplification, and a good one. But still, when rounding 1.5, both 1 and 2 are just as accurate choices, mathematically; it's the exact same difference.
Indeed, there are other methods, but as I said, they are irrelevant, and your argument shiftless. We have accepted standards such that the only time you should need to explain what you are doing is when you are violating those standard. No one mentioned which way the rounding was going when the question was posed, therefore we fall back on the accepted norm of rounding up for exact mid points. In fact, given the context and background information of the project, it becomes obvious that that was the intent. Your pedantry is incorrectly applied in this case.
 
Real life tends to teach me otherwise although I don´t approve much to it when it comes to ethics...


It goes like this:


1.5 = 1 if I have to pay


1.5 = 2 if I have to receive


:p :p :p
 
jeeeeesus fucking christ! its Two Months! okay? the guy will get his damn pandora in Two Months......FUCK!
 
Real life tends to teach me otherwise although I don´t approve much to it when it comes to ethics...


It goes like this:


1.5 = 1 if I have to pay


1.5 = 2 if I have to receive


:p :p :p
What if the thing you are paying for & receiving is a Pandora, and the number refers to the number of Months wait?
 
Oh god, then I have no idea on how to conjecture that... but you have to take on account the object delivered is great and priceless!


My example was more like how the banks deal numbers with me, and I was wrong, considering banks here in brazil, it is more like this:


1.5 = 1.1 if I lend them money


1.5 = 4.5 if I borrow from them


within 1 year.
 
Like jkt mentions, there are other methods of rounding. I conjecture that we generally use the "round up" method (well, actually, the "round away from 0" method) because it means it is possible to accurately round any floating-point number with more digits than the tenths place while only looking at the tenths place; 0.50001 is technically closer to 1 than 0, so rounding up is more accurate than rounding down. On the other hand, if we used the "round to even" rule, we would have to always remember to look past the tenths place and always round up if it isn't exactly *.5, because if we rounded 2.52 to 2, that would be less accurate than rounding it to 3. So rounding up always is a simplification, and a good one. But still, when rounding 1.5, both 1 and 2 are just as accurate choices, mathematically; it's the exact same difference.
Indeed, there are other methods, but as I said, they are irrelevant, and your argument shiftless. We have accepted standards such that the only time you should need to explain what you are doing is when you are violating those standard. No one mentioned which way the rounding was going when the question was posed, therefore we fall back on the accepted norm of rounding up for exact mid points. In fact, given the context and background information of the project, it becomes obvious that that was the intent. Your pedantry is incorrectly applied in this case.
Here's what I think: when rounding 1.5 months to either 1 or 2 months, you'd better have justification, because you're either adding or subtracting 1/3 of the exact estimate! That's pretty significant! But if you must round, it hardly matters which way. 1 or 2 months, it's the same difference. Either way is 0.5 months difference, or approximately 15 days out of approximately 45. I may have to justify why it's going to be 15 days earlier, but you would just as well have to justify why it's going to be 15 days later. The only reason for justification being needed is it's a huge difference.
 
Here's what I think: when rounding 1.5 months to either 1 or 2 months, you'd better have justification, because you're either adding or subtracting 1/3 of the exact estimate! That's pretty significant!
We're talking about Pandora delivery here, right? Given the history of the project, I think rounding 1.5 up to 2 months is pretty conservative :p
 
Back
Top