Submit | All submissions | Best solutions | Back to list |
THREETW1 - Connecting three towns (variation) |
Given three towns find the point for which the summary distance to all towns is minimal. Wanted are the accurate coordinates ±10-6.
Input
In the first line the number T (T<100) of test cases. Then T lines with the space separated coordinates (integer 0<=x, y<=100) of the three towns.
Output
For each test case the coordinates of the above specified point.
Example
Input: 2
2 2 1 1 2 1
1 0 0 0 2 1 Output: 1.788675 1.211324
1 0
Added by: | HWK |
Date: | 2011-08-25 |
Time limit: | 0.100s |
Source limit: | 50000B |
Memory limit: | 1536MB |
Cluster: | Cube (Intel G860) |
Languages: | All except: ASM64 |
hide comments
|
|||||
2011-08-29 11:36:49 HWK
@Aurelian Tutuianu: I use the standard SPOJ judge FP 10^-6. That's its comparing method: static bool equal (lexem a, lexem b) { if (a.type != b.type) return false; if (a.type==TP_EOF) return true; if (a.type==TP_CHAR) return (a.c==b.c); return (fabs(a.d-b.d)<=epsilon); } with double epsilon=0.000001; This method compares your output with my results which I've saved with 10 digits after the decimal point. Except for choosing the judge and preparing the test data I have no influence on the judging. I didn't find a better way to specify the 'rounding rules' of this judge than '+-10^-6'. That's one error in your output: input - 16 75 27 13 56 41 your output - 39.942763 39.811774 my ouput - 39.9427649493 39.8117728195 I hope this helps to solve your problem. Last edit: 2011-08-29 11:16:32 |
|||||
2011-08-29 11:36:49 Aurelian Tutuianu
@hwk: I consider that the problem should consist only in algorithm, not in a silly formatting thing. Especially when the formatting rules are not specified. I don't want to waste time to guess what rounding to apply. Perhaps you should better specify rounding rules in the problem statement. I renounced to solve it, even if I know the solution and the problem is nice. |
|||||
2011-08-29 11:36:49 Damian Straszak
What is the answer for 0 0 1 0 5 0? There is no "intersection" in the optimal path network. You probably mean: find the point for which the summary distance to all three towns is minimal. And tell me please why: "'%.6f' isn't enough for AC."? |
|||||
2011-08-29 11:36:49 HWK
@Pranay: Look at the last comment. You have to minimize the sum of the lengths. And keep in mind: '%.6f' isn't enough for AC. Hi, Aurelian Tutuianu! Look at your output! The precision isn't enough. ;-) Last edit: 2011-08-27 20:16:26 |
|||||
2011-08-29 11:36:49 Pranay
@hwk: is the sum of the lenghts minimum or each individual length? nice problem :) Last edit: 2011-09-02 02:13:48 |
|||||
2011-08-29 11:36:49 HWK
@Oleg: Why have you removed your question and my answer? Here's a recapitulation of the two deleted comments: 1. You have to minimize the summary length of all streets. 2. '%.6f' isn't enough for AC. Last edit: 2011-08-27 08:09:24 |