|
|
вернуться в форумPlease check the 12-th test. I and others thinks that our solutions are correct, but they are WA on 12... Maybe anyone can give a little hint? Edited by author 03.08.2006 16:49 Thank you! :) But it is not helped me. Maybe you give me this test? Or any similar? Yeah, I have WA12 too. I doubt if the test is really correct. And one more question: what means pluses and minuses in brackets? (Sorry for offtop). Edited by author 03.08.2006 22:16 Edited by author 03.08.2006 22:16 Cool men write (-) if there are only theme in post, and (+) if it consists any more than ony theme :) It seems that 12 test has cases like 1e14234320000000 2 Is that output would be correct for this case: 1.00e14234320000000 Edited by author 04.08.2006 01:42 12 test has cases like 1e-14234320000000 2 and correct answer for this case: 0.00 From my code: if( exp > 1000 ) ... // make Output Limit The result is Output Limit, test 12! So, exp is more than 1000 :) From problem statement: "It is guaranteed that a length of a result will not exceed 200 symbols." Is it right (with such big exp and number != 0)? No, that was bug in my check code. Later I submited smth like that: if (e>1000 && !isnull(mantissa)) while (1); and this code didn't cause tle. I had the code: if( abs( exp ) > 1000 ) ... // the answer is 0.0000... It got WA 12. Then I decided to divide it into two parts: exp>1000 and exp<-1000. if( exp < -1000 ) ... // 0.000... - 100%, yes? else if( exp > 1000 ) { // I amn't sure that there can be such a situation, that's why I do only output limit while( 1 ) cout << "111\n"; } Output Limit 12! "It is guaranteed the the length of output would be less than 200 symbols". I think, exp>1000 can be only when the number is 0 ( 0e10000 ). But then the right answer is 0.000 - the first code, which gets WA. So, there is a test in 12th test where exp>1000 and number is NOT 0! You forget that input file may contain a number of test cases. Maybe cases with big (and small) exponent your program processes correctly, but fails at other cases. Although i agree that it is very, very strange. There are tests like 1e+00000000000000001 5 or 1e-10000000000000000 2 First, my program had WA on test 12. Than I fixed my bug with this tests, and finally got AC=) my program processes correctly these cases, outputing: 10.00000 0.00 but still wa#12 Try this test: 1+10 0 # it's not a correct float number |
|
|