Sorry your browser is not supported!

You are using an outdated browser that does not support modern web technologies, in order to use this site please update to a new browser.

Browsers supported include Chrome, FireFox, Safari, Opera, Internet Explorer 10+ or Microsoft Edge.

DarkBASIC Professional Discussion / Strange calc error

Author
Message
Fane
14
Years of Service
User Offline
Joined: 9th May 2012
Location:
Posted: 28th Aug 2014 23:00
Link
Hello everybody its me again
I have here a strange calculation error.



vel# should be always 20 not matter how many frames I get. It works with 60 frames and above but at 45 frames I get the first small errors like 22,5 and it goes up to 30 with 30 frames O.o

But I am sure vel# = 30 / (30 /20) = 20 O.o

Someone know what cause this strange problem?

System: Win 8.1 64 / Core i5 (3,9GHz), GeForce 760 GTX 192 - Bit Dx 11 (Driver 335.23) / 8GB Ram
Back to top
Profile PM
Ortu
DBPro Master
18
Years of Service
User Offline
Joined: 21st Nov 2007
Location: Austin, TX
Posted: 29th Aug 2014 01:02 Edited at: 29th Aug 2014 01:06
Link
Floats have a measure of imprecision to start with, mixing floats and integers will only make things murkier.

If you want vel# to always be 20 regardless of fps, why even do a calculation?

You can also reduce to one function call to screen fps() by just setting it to start with



Back to top
Profile PM Email Website
Green Gandalf
VIP Member
21
Years of Service
User Offline
Joined: 3rd Jan 2005
Playing: Malevolence:Sword of Ahkranox, Skyrim, Civ6.
Posted: 29th Aug 2014 01:33
Link
More precisely, the problem is that your expression uses integer arithmetic so 30/(30/20) becomes 30/1 (because the real number 1.5 (=30/20) gets rounded down to the integer 1) and then 30/1 evaluates to 30.

If you want more exact division using floats (which as Ortu suggest only have a few significant figures accuracy) then you could do something like



That would force the more accurate float division throughout because the first expression evaluated is fps/20.0 and the presence of a float forces conversion to a float which in turn forces the next division to be a float as well.

But why would you want that anyway? As Ortu says, why not simply write



etc?



Powered by Free Banners
Back to top
Profile PM Email
Fane
14
Years of Service
User Offline
Joined: 9th May 2012
Location:
Posted: 29th Aug 2014 09:01
Link
Thanks guys but after 12 hours of programming yesterday I dont get this simple idea anymore

@Green Gandalf: Good to know with vel# = fps / (fps / 20.0)

System: Win 8.1 64 / Core i5 (3,9GHz), GeForce 760 GTX 192 - Bit Dx 11 (Driver 335.23) / 8GB Ram
Back to top
Profile PM
Libervurto
20
Years of Service
User Offline
Joined: 30th Jun 2006
Location: On Toast
Posted: 31st Aug 2014 11:53 Edited at: 31st Aug 2014 12:07
Link
Were you drunk when you wrote this or are you just new? Either way you will be laughing at this code in a week.

I agree with Ortu's code but I would also store vel# = 20.0 as a constant, in case you want to change its value in future.

Quote: "Thanks guys but after 12 hours of programming yesterday I dont get this simple idea anymore"

It is like calculating 6 + 1 - 1 = 6, it is a waste of time because you are just cancelling out the calculation. This might not be so obvious with division... [goes off to make a latex example]

You should remember this from algebra



Formerly OBese87.
Back to top
Profile PM Email
Ortu
DBPro Master
18
Years of Service
User Offline
Joined: 21st Nov 2007
Location: Austin, TX
Posted: 6th Sep 2014 05:50
Link
Quote: " I would also store vel# = 20.0 as a constant, in case you want to change its value in future."


good point

Back to top
Profile PM Email Website
Back to top

Login to post a reply

Server time is: 2026-07-05 17:32:54
Your offset time is: 2026-07-05 17:32:54