Problem 178: Step Numbers

(see projecteuler.net/problem=178)

Consider the number 45656.
It can be seen that each pair of consecutive digits of 45656 has a difference of one.
A number for which every pair of consecutive digits has a difference of one is called a step number.
A pandigital number contains every decimal digit from 0 to 9 at least once.
How many pandigital step numbers less than 10^40 are there?

My Algorithm

And again: it's time for Dynamic Programming ...
My search function has three parameters:
1. a mask where bit x is set if any of the digits in front of the current digit was x.
2. the current digit
3. how many digits still have to be processed

A number is pandigital if the all bits 0..9 are set (AllDigits = (1 << 10) - 1 which is 1023).
If the current digit is bigger than 0, then the next can be "a step down".
If the current digit is smaller than 9, then the next can be "a step up".

Of course, this algorithm would be far too slow, but the algorithm can heavily memoize results.
The function's parameters are merged into a single unique ID and stored in a cache (with 1024*40*10 = 409600 entries, which requires about 3 MByte RAM).

Modifications by HackerRank

They want you to compute the result for n < 10^{10^4}. I can't figure out whether their input is always a power-of-ten or not.
The result will easily be too large for C++'s 64 bit integers anyway.

Note

If the first digit is between 1 and 4 (inclusive), then for every solution there is a "mirrored" version, where each digit b_i = 9 - a_i.
For example, 10123456789 is a solution and therefore 89876543210 is a solution, too.
How does it help ? Well, you can compute all solutions starting with 1..4, multiply that number by 2 and add all solutions starting with 9.

Interactive test

You can submit your own input to my program and it will be instantly processed at my server:

Input data (separated by spaces or newlines):

This is equivalent to
echo 11 | ./178

Output:

(please click 'Go !')

Note: the original problem's input 40 cannot be entered
because just copying results is a soft skill reserved for idiots.

(this interactive test is still under development, computations will be aborted after one second)

My code

… was written in C++ and can be compiled with G++, Clang++, Visual C++. You can download it, too.

#include <iostream>
#include <vector>
 
// initial state: no digits used, all bits are zero
const unsigned int NoDigits = 0;
// all ten digits used => all ten bits set
const unsigned int AllDigits = (1 << 10) - 1; // = 1023
// analyze all numbers with up to 10^40
const unsigned int MaxDigits = 40;
 
// count pandigital step number where bits in "mask" are set according to the first digits
// and the current digit is "currentDigit" with "numDigitsLeft" digits to go
unsigned long long search(unsigned int mask, unsigned int currentDigit, unsigned int numDigitsLeft)
{
// update mask, "use digit"
mask |= 1 << currentDigit;
 
// arrived at last digit ?
if (numDigitsLeft == 1)
{
// all ten bits set ? => pandigital
if (mask == AllDigits)
return 1; // yes, a solution
 
// not pandigital, no valid solution
return 0;
}
 
// memoize
unsigned int hash = mask * MaxDigits * 10 + (numDigitsLeft - 1) * 10 + currentDigit;
static std::vector<unsigned long long> cache(1024 * 10 * MaxDigits, 0);
// result already known ?
if (cache[hash] != 0)
return cache[hash];
 
unsigned long long result = 0;
// next digit is smaller
if (currentDigit > 0)
result += search(mask, currentDigit - 1, numDigitsLeft - 1);
// next digit is bigger
if (currentDigit < 9)
result += search(mask, currentDigit + 1, numDigitsLeft - 1);
 
cache[hash] = result;
return result;
}
 
int main()
{
// maxDigits is the actual number of digits
// MaxDigits is the highest allowed value for maxDigits
// I know, it's confusing ...
unsigned int maxDigits = 40;
std::cin >> maxDigits;
// note: input syntax differs from Hackerrank !
 
unsigned long long result = 0;
// for each number of digits ...
for (unsigned int numDigits = 1; numDigits <= maxDigits; numDigits++)
{
// ... the first one must be 1..9 (and never zero !)
for (unsigned int digit = 1; digit <= 9; digit++)
result += search(NoDigits, digit, numDigits);
 
// a little bit faster:
//result += 2 * search(NoDigits, 1, numDigits);
//result += 2 * search(NoDigits, 2, numDigits);
//result += 2 * search(NoDigits, 3, numDigits);
//result += 2 * search(NoDigits, 4, numDigits);
//result += search(NoDigits, 9, numDigits);
}
 
std::cout << result << std::endl;
return 0;
}

This solution contains 11 empty lines, 25 comments and 2 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.
Peak memory usage was about 5 MByte.

(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

June 28, 2017 submitted solution
June 28, 2017 added comments

Hackerrank

see https://www.hackerrank.com/contests/projecteuler/challenges/euler178

My code solves 2 out of 25 test cases (score: 4.17%)

I failed 23 test cases due to wrong answers and 0 because of timeouts

Difficulty

55% Project Euler ranks this problem at 55% (out of 100%).

Hackerrank describes this problem as medium.

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
  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
The 289 solved problems (level 11) had an average difficulty of 32.1% at Project Euler and
I scored 13,486 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.

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 !