<< problem 106 - Special subset sums: meta-testing Diophantine reciprocals I - problem 108 >>

# Problem 107: Minimal network

The following undirected network consists of seven vertices and twelve edges with a total weight of 243.

The same network can be represented by the matrix below.

ABCDEFG
A 161221
B16  1720
C12  28 31
D211728 181923
E 20 18  11
F  3119  27
G   231127

However, it is possible to optimise the network by removing some edges and still ensure that all points on the network remain connected.
The network which achieves the maximum saving is shown below. It has a weight of 93, representing a saving of 243 - 93 = 150 from the original network.

Using network.txt (right click and 'Save Link/Target As...'), a 6K text file containing a network with forty vertices,
and given in matrix form, find the maximum saving which can be achieved by removing redundant edges whilst ensuring that the network remains connected.

# My Algorithm

I use Prim's algorithm (see en.wikipedia.org/wiki/Prim's_algorithm):

• done contains these nodes where all connections are optimized
• done contains initially an arbitrarily chosen node
• algorithm is finished when done contains all nodes
• in each iteration, one node is added to done
• store all edges in next where one node is in done and the other node isn't
• sort next by the weight
• pick the lowest
• add that edge to the optimized graph minimal
• add the other edge's node to done, too
There will be 40-1=39 iterations. The most time-consuming part is locating all potential edges.
The "true" algorithm of Prim uses a priority queue quite efficiently whereas I rebuild it from scratch every time.
Nevertheless, my code solves the 2500x2500 networks of Hackerrank within less than a second.

## Modifications by HackerRank

• the input format is completely different
• two nodes may be connected via multiple edges
• print the optimized weight instead of the gain

# Interactive test

This feature is not available for the current problem.

# My code

… was written in C++11 and can be compiled with G++, Clang++, Visual C++. You can download it, as well as the input data, too.

The code contains #ifdefs to switch between the original problem and the Hackerrank version.
Enable #ifdef ORIGINAL to produce the result for the original problem (default setting for most problems).

       //#define ORIGINAL

#include <iostream>

#include <set>
#include <map>
#include <queue>

typedef unsigned int Node;
typedef unsigned int Weight;

// connect two nodes and assign a weight
struct Edge
{
Node from;
Node to;
Weight weight;

// for sorting
bool operator<(const Edge& other) const
{
// STL's priority queue returns the __largest__ element first, therefore I invert the comparison sign
if (weight != other.weight)
return weight > other.weight;

// sort by nodes, too, to avoid collisions of nodes with the same weight
// the order doesn't matter, these nodes just "have to be different"
if (from   != other.from)
return from   < other.from;

}
};

// all edges (full network)
std::map<std::pair<Node, Node>, Weight> network;
// all nodes
std::set<Node> nodes;
// weight of the full network, including duplicate connections
unsigned int initialSum = 0;

// insert an edge into the network, true if already existing
void addEdge(Node from, Node to, Weight weight)
{
// count all originals weights
initialSum += weight;

// all edges have the lower ID first (I can do that because it's an undirected graph)
if (from > to)
std::swap(from, to);

auto fromTo = std::make_pair(from, to);
if (network.count(fromTo) != 0)
// has the existing edge a lower or the same weight ? => we're done
if (network[fromTo] <= weight)
return;

// store new (or improved) connection between two nodes
network[fromTo] = weight;
nodes.insert(from);
nodes.insert(to);
}

int main()
{
#ifdef ORIGINAL
// read Project Euler's weird file format
for (unsigned int i = 0; i < 40; i++)
for (unsigned int j = 0; j < 40; j++)
{
char c = 0;
unsigned int weight = 0;
// CSV-format
while (c != ',' && c != '\n')
{
c = std::cin.get();
// ignore dashes
if (c >= '0' && c <= '9')
{
weight *= 10;
weight += c - '0';
}
}

// only valid weights
if (weight != 0 && i < j)
}

#else

unsigned int numNodes, numEdges;
std::cin >> numNodes >> numEdges;
for (unsigned int i = 0; i < numEdges; i++)
{
Node   from, to;
Weight weight;
std::cin >> from >> to >> weight;
}
#endif

// optimized graph, initially empty
std::set<Edge> minimal;

std::set<Node> done;
done.insert(*nodes.begin());

// not all nodes optimized yet ?
while (done.size() < nodes.size())
{
// add all edges where one node is part of tree and the other isn't
std::priority_queue<Edge> next;
for (auto e : network)
{
auto fromTo  = e.first;
bool hasFrom = done.count(fromTo.first)  != 0;
bool hasTo   = done.count(fromTo.second) != 0;
if (hasFrom == hasTo) // both nodes are already optimized or both aren't ?
continue;

next.push({fromTo.first, fromTo.second, e.second});
}

// get edge with minimal weight

// add new edge to tree
done.insert(add.from); // one insert is redundant but set::set throws it away

// add edge to the optimized graph
}

// "measure" both graphs
unsigned int optimizedSum = 0;
for (auto i : minimal)
optimizedSum += i.weight;

#ifdef ORIGINAL
// difference
auto gain = initialSum - optimizedSum;
std::cout << gain << std::endl;
#else
std::cout << optimizedSum << std::endl;
#endif

return 0;
}


This solution contains 25 empty lines, 28 comments and 10 preprocessor commands.

# Benchmark

The correct solution to the original Project Euler problem was found in less than 0.01 seconds on an Intel® Core™ i7-2600K CPU @ 3.40GHz.
(compiled for x86_64 / Linux, GCC flags: -O3 -march=native -fno-exceptions -fno-rtti -std=gnu++11 -DORIGINAL)

See here for a comparison of all solutions.

Note: interactive tests run on a weaker (=slower) computer. Some interactive tests are compiled without -DORIGINAL.

# Changelog

May 11, 2017 submitted solution

# Hackerrank

My code solves 8 out of 8 test cases (score: 100%)

# Difficulty

Project Euler ranks this problem at 35% (out of 100%).

Hackerrank describes this problem as easy.

Note:
Hackerrank has strict execution time limits (typically 2 seconds for C++ code) and often a much wider input range than the original problem.
In my opinion, Hackerrank's modified problems are usually a lot harder to solve. As a rule thumb: brute-force is rarely an option.

# Heatmap

Please click on a problem's number to open my solution to that problem:

 green solutions solve the original Project Euler problem and have a perfect score of 100% at Hackerrank, too yellow solutions score less than 100% at Hackerrank (but still solve the original problem easily) gray problems are already solved but I haven't published my solution yet blue solutions are relevant for Project Euler only: there wasn't a Hackerrank version of it (at the time I solved it) or it differed too much orange problems are solved but exceed the time limit of one minute or the memory limit of 256 MByte red problems are not solved yet but I wrote a simulation to approximate the result or verified at least the given example - usually I sketched a few ideas, too black problems are solved but access to the solution is blocked for a few days until the next problem is published [new] the flashing problem is the one I solved most recently
 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100
 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200
 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300
 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400
 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500
 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600
 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633
The 310 solved problems (that's level 12) had an average difficulty of 32.6% at Project Euler and
I scored 13526 points (out of 15700 possible points, top rank was 17 out of ≈60000 in August 2017) at Hackerrank's Project Euler+.

My username at Project Euler is stephanbrumme while it's stbrumme at Hackerrank.

Look at my progress and performance pages to get more details.

 << problem 106 - Special subset sums: meta-testing Diophantine reciprocals I - problem 108 >>
more about me can be found on my homepage, especially in my coding blog.
some names mentioned on this site may be trademarks of their respective owners.
thanks to the KaTeX team for their great typesetting library !