mersenneforum.org Minimal set of the strings for primes with at least two digits
 Register FAQ Search Today's Posts Mark Forums Read

2021-10-05, 12:07   #188
sweety439

"99(4^34019)99 palind"
Nov 2016
(P^81993)SZ base 36

1100001011012 Posts

Quote:
 Originally Posted by sweety439 Some known unsolved families for bases b<=64 not in the list for bases 2 to 30 or list for bases 28 to 50: Base 11: 5{7} (found by me) Base 13: 9{5} (found by me) Base 13: A{3}A (found by me) Base 17: 15{0}D (found by me) Base 17: 1F{0}7 (found by me) Base 18: C{0}C5 (found by me) Base 25: F{2} (found by extended generalized Riesel conjecture base 25 with k > CK) Base 31: 2{F} (found by extended generalized Riesel conjecture base 31) Base 31: 3{5} (found by extended generalized Riesel conjecture base 31) Base 32: 8{0}V (see https://oeis.org/A247952) Base 32: S{V} (found by CRUS generalized Riesel conjecture base 1024) Base 37: 2K{0}1 (found by CRUS generalized Sierpinski conjecture base 37) Base 37: {I}J (see http://www.fermatquotient.com/PrimSerien/GenFermOdd.txt) Base 38: 1{0}V (see https://math.stackexchange.com/quest...the-form-38n31) Base 43: 2{7} (found by extended generalized Riesel conjecture base 43) Base 43: 3b{0}1 (found by CRUS generalized Sierpinski conjecture base 43) Base 53: 19{0}1 (found by CRUS generalized Sierpinski conjecture base 53) Base 53: 4{0}1 (found by CRUS generalized Sierpinski conjecture base 53) Base 55: a{0}1 (found by CRUS generalized Sierpinski conjecture base 55) Base 55: {R}S (see http://www.fermatquotient.com/PrimSerien/GenFermOdd.txt) Base 60: Z{x} (see CRUS generalized Riesel conjecture base 60) Base 62: 1{0}1 (see http://jeppesn.dk/generalized-fermat.html) Base 63: {V}W (see http://www.fermatquotient.com/PrimSerien/GenFermOdd.txt)
No, base 32 family 8{0}V has covering set {3, 5, 41}, thus can be ruled out as only contain composite numbers.

 2021-10-06, 23:52 #189 sweety439     "99(4^34019)99 palind" Nov 2016 (P^81993)SZ base 36 1100001011012 Posts New minimal prime (start with b+1) in base b is found for b=908: 8(0^243438)1, see post https://mersenneforum.org/showpost.p...&postcount=992 File https://docs.google.com/spreadsheets...RwmKME/pubhtml updated.
 2021-10-16, 13:38 #190 sweety439     "99(4^34019)99 palind" Nov 2016 (P^81993)SZ base 36 60558 Posts Conjecture: If sequence (a*b^n+c)/gcd(a+c,b-1) (a>=1 is integer, b>=2 is integer, c is (positive or negative) integer, |c|>=1, gcd(a,c) = 1, gcd(b,c) = 1) does not have covering set (full numerical covering set, full algebraic covering set, or partial algebraic/partial numerical covering set), then the sum of the reciprocals of the positive integers n such that (a*b^n+c)/gcd(a+c,b-1) is prime is converge (i.e. not infinity) and transcendental number. (of course, this conjecture will imply that there are infinitely many such n) For the examples of (a,b,c) triples (a>=1 is integer, b>=2 is integer, c is (positive or negative) integer, |c|>=1, gcd(a,c) = 1, gcd(b,c) = 1) such that (a*b^n+c)/gcd(a+c,b-1) have covering set (full numerical covering set, full algebraic covering set, or partial algebraic/partial numerical covering set), see post https://mersenneforum.org/showpost.p...&postcount=678
 2021-10-16, 13:45 #191 sweety439     "99(4^34019)99 palind" Nov 2016 (P^81993)SZ base 36 60558 Posts Another conjecture (seems to already be proven, but I am not sure that): If all but finitely many primes p divide (a*b^n+c)/gcd(a+c,b-1) (a>=1 is integer, b>=2 is integer, c is (positive or negative) integer, |c|>=1, gcd(a,c) = 1, gcd(b,c) = 1) for some n>=1, then a=1 and c=-1, i.e. (a*b^n+c)/gcd(a+c,b-1) is generalized repunit number (b^n-1)/(b-1) The factor tables have many examples for the special case that b=10, e.g. {2}1 in base 10 is (a,b,c) = (2,10,-11), the section Prime factors that appear periodically lists the primes that divide (a*b^n+c)/gcd(a+c,b-1) = (2*10^n-11)/9 for some n, and we note that the primes 2, 5, 11, 31, 37, 41, 43, 53, 71, 73, 79, 83, 101, 103, 107, 127, 137, 157, 173, 191, 199, 227, 239, 241, 251, 271, 281, 283, 307, 311, 317, 331, 347, 349, 353, 397, 409, 449, 523, 547, 563, 569, 599, 601, 613, 617, 631, 641, 643, 653, 661, 673, 691, 719, 733, 739, 751, 757, 761, 769, 773, 787, 797, 809, 827, 829, 839, 853, 859, 907, 911, 967, 991, 997, ..., divides no numbers of the form (a*b^n+c)/gcd(a+c,b-1) = (2*10^n-11)/9, and this sequence of primes seems to be infinite, another example is 1{3} in base 10 is (a,b,c) = (4,10,-1), the section Prime factors that appear periodically lists the primes that divide (a*b^n+c)/gcd(a+c,b-1) = (4*10^n-1)/3 for some n, and we note that the primes 2, 3, 5, 11, 37, 41, 53, 73, 79, 101, 103, 137, 139, 173, 211, 239, 241, 271, 277, 281, 317, 331, 349, 353, 397, 421, 449, 463, 521, 547, 607, 613, 617, 661, 673, 733, 751, 757, 773, 797, 829, 853, 859, 907, 967, ..., divides no numbers of the form (a*b^n+c)/gcd(a+c,b-1) = (4*10^n-1)/3, and this sequence of primes seems to be infinite. Of course this conjecture also include for bases other than 10, i.e. for every (a*b^n+c)/gcd(a+c,b-1) (a>=1 is integer, b>=2 is integer, c is (positive or negative) integer, |c|>=1, gcd(a,c) = 1, gcd(b,c) = 1) family other than generalized repunit family (b^n-1)/(b-1) (i.e. a=1 and c=-1), there are infinitely many primes not dividing any number of the form (a*b^n+c)/gcd(a+c,b-1), e.g. for the base 11 unsolved family 5{7} = (57*11^n-7)/10, the primes dividing no numbers of the form (a*b^n+c)/gcd(a+c,b-1) = (57*11^n-7)/10 are 3, 7, 11, 19, 37, 43, 61, 83, 89, 107, 131, 137, 157, 191, 193, 199, 211, 229, 241, 257, 269, 307, 311, 313, 317, 379, 389, 397, 421, 431, 439, 449, 457, 479, 503, 509, 521, 523, 541, 547, 571, 577, 607, 617, 631, 641, 653, 659, 661, 691, 727, 739, 743, 751, 757, 773, 787, 797, 811, 827, 829, 907, 911, 919, 967, ..., and for the base 17 unsolved family F1{9} = (4105*17^n-9)/16, the primes dividing no numbers of the form (a*b^n+c)/gcd(a+c,b-1) = (4105*17^n-9)/16 are 3, 5, 17, 29, 43, 59, 67, 71, 79, 101, 103, 137, 151, 157, 163, 179, 181, 191, 199, 223, 229, 239, 241, 257, 263, 281, 293, 307, 331, 337, 353, 359, 373, 383, 389, 409, 433, 443, 457, 461, 463, 491, 509, 541, 563, 587, 601, 619, 631, 647, 659, 661, 727, 733, 739, 757, 761, 769, 773, 797, 811, 821, 829, 859, 863, 877, 883, 919, 937, 947, 953, 967, 977, 991, ... (it is surprising that many primes do not divide (4105*17^n-9)/16 for any n, since (4105*17^n-9)/16 is a low-weight form, i.e. (4105*17^n-9)/16 is divisible by a small prime for most n, note that (4105*17^n-9)/16 has no algebraic factorization for any n, since 4105 is not perfect power) Last fiddled with by sweety439 on 2021-10-18 at 14:46
 2021-10-16, 13:54 #192 sweety439     "99(4^34019)99 palind" Nov 2016 (P^81993)SZ base 36 3×1,039 Posts The smallest generalized near-repdigit primes (i.e. of the form x{y} or {x}y) base b is always minimal primes (start with b+1) in base b unless the repeating digit (i.e. y for x{y}, or x for {x}y) is 1, since the generalized repunit numbers base b may be prime unless b is 9, 25, 32, 49, 64, 81, 121, 125, 144, ... (A096059) bases without any generalized repunit primes, and for a repdigit to be prime, it must be a repunit (i.e. the repeating digit is 1) and have a prime number of digits in its base (except trivial single-digit numbers), since, for example, the repdigit 77777 is divisible by 7, in any base > 7, thus (i.e. of the form x{y} or {x}y) base b is always minimal primes (start with b+1) in base b if the repeating digit (i.e. y for x{y}, or x for {x}y) is not 1, thus, the families A{1} in base 22 and 8{1} in base 33 and 4{1} in base 40 are not unsolved families in this problem (i.e. finding all minimal primes (start with b+1) in base b) although all they are near-repdigit families and all they have no known primes or PRPs and none of them can be ruled out as only contain composites (only count numbers > base), since their repeating digit are 1, and the prime F(1^957) in base 24 (its value is (346*24^957-1)/23) is not minimal prime (start with b+1) in base b=24, since its repeating digit is 1 Last fiddled with by sweety439 on 2021-10-17 at 12:55
 2021-10-17, 17:00 #193 sweety439     "99(4^34019)99 palind" Nov 2016 (P^81993)SZ base 36 3×1,039 Posts The algebra form ((a*b^n+c)/d) of the unsolved families are: Code: base unsolved family algebra form 11 5(7^n) (57*11^n-7)/10 13 9(5^n) (113*13^n-5)/12 13 A(3^n)A (41*13^(n+1)+27)/4 16 (3^n)AF (16^(n+2)+619)/5 16 (4^n)DD (4*16^(n+2)+2291)/15 See https://stdkmd.net/nrr/exprgen.htm for the algebra form calculator (only for base 10 families), also see page 16 of https://cs.uwaterloo.ca/~cbright/reports/mepn.pdf
 2021-10-18, 18:04 #194 sweety439     "99(4^34019)99 palind" Nov 2016 (P^81993)SZ base 36 3·1,039 Posts Families which can be ruled out as contain no primes (only count numbers > base) by reasons other than trivial 1-cover are: Code: Base 5: {1}3 (covering set {2,3}) (not produce minimal primes (start with b+1) since 111 is prime) {1}4 (covering set {2,3}) (not produce minimal primes (start with b+1) since 111 is prime) 3{1} (covering set {2,3}) (not produce minimal primes (start with b+1) since 111 is prime) 4{1} (covering set {2,3}) (not produce minimal primes (start with b+1) since 111 is prime) Base 8: 1{0}1 (sum of cubes) 6{4}7 (covering set {3,5,13}) (not produce minimal primes (start with b+1) since 44444444444444444444444444444444444444444444444444444444444444444444444444444444444444444444444444444444444444444444444444444444444444444444444444444444444444444444444444444444444444444444444444444444444444444444444444447 is prime) Base 9: {1} (difference of squares) {1}5 (covering set {2,5}) 2{7} (covering set {2,5}) 3{1} (difference of squares) {3}5 (covering set {2,5}) 3{8} (difference of squares) {3}8 (covering set {2,5}) 5{1} (covering set {2,5}) 5{7} (covering set {2,5}) 6{1} (covering set {2,5}) {7}2 (covering set {2,5}) {7}5 (covering set {2,5}) {8}5 (difference of squares) Base 11: {1}3 (covering set {2,3}) (not produce minimal primes (start with b+1) since 11111111111111111 is prime) {1}4 (covering set {2,3}) (not produce minimal primes (start with b+1) since 11111111111111111 is prime) {1}9 (covering set {2,3}) (not produce minimal primes (start with b+1) since 11111111111111111 is prime) {1}A (covering set {2,3}) (not produce minimal primes (start with b+1) since 11111111111111111 is prime) 2{5} (covering set {2,3}) 3{1} (covering set {2,3}) (not produce minimal primes (start with b+1) since 11111111111111111 is prime) 3{5} (covering set {2,3}) 3{7} (covering set {2,3}) 4{1} (covering set {2,3}) (not produce minimal primes (start with b+1) since 11111111111111111 is prime) 4{7} (covering set {2,3}) {5}2 (covering set {2,3}) {5}3 (covering set {2,3}) {5}8 (covering set {2,3}) {5}9 (covering set {2,3}) {7}4 (covering set {2,3}) {7}9 (covering set {2,3}) {7}A (covering set {2,3}) 8{5} (covering set {2,3}) 9{1} (covering set {2,3}) (not produce minimal primes (start with b+1) since 11111111111111111 is prime) 9{5} (covering set {2,3}) 9{7} (covering set {2,3}) A{1} (covering set {2,3}) (not produce minimal primes (start with b+1) since 11111111111111111 is prime) A{7} (covering set {2,3}) Base 12: {B}9B (combined with covering set {13} and difference of squares) Last fiddled with by sweety439 on 2021-10-26 at 20:56
 2021-10-19, 11:04 #195 sweety439     "99(4^34019)99 palind" Nov 2016 (P^81993)SZ base 36 3·1,039 Posts Like http://www.wiskundemeisjes.nl/wp-con...02/primes2.pdf, if you write down a prime > 10, then I can always strike out 0 or more digits to get a prime on this set: {11, 13, 17, 19, 23, 29, 31, 37, 41, 43, 47, 53, 59, 61, 67, 71, 73, 79, 83, 89, 97, 227, 251, 257, 277, 281, 349, 409, 449, 499, 521, 557, 577, 587, 727, 757, 787, 821, 827, 857, 877, 881, 887, 991, 2087, 2221, 5051, 5081, 5501, 5581, 5801, 5851, 6469, 6949, 8501, 9001, 9049, 9221, 9551, 9649, 9851, 9949, 20021, 20201, 50207, 60649, 80051, 666649, 946669, 5200007, 22000001, 60000049, 66000049, 66600049, 80555551, 555555555551, 5000000000000000000000000000027} In fact, if a primes which do not contain any of the 1st to the 76th primes in the set, then this prime must be of the form 5{0}27 = 5*10^(n+2)+27, and if a primes which do not contain any of the 1st to the 75th primes in the set, then the prime must be of one of these forms: 5{0}27 = 5*10^(n+2)+27, {5}1 = (5*10^(n+1)-41)/9, 8{5}1 = (77*10^(n+1)-41)/9, by our theorem, such primes must contain either the 76th prime or the 77th prime as subsequence, however any such prime cannot contain both the 76th prime and the 77th prime as subsequences, since the 76th prime contain "1" and the 77th prime contain "7", any prime (in fact, any number, need not to be prime) containing both the 76th prime and the 77th prime as subsequences will contain both "1" and "7", and hence contain either "17" or "71" (or both) as subsequence, but both 17 and 71 are primes > 10 Last fiddled with by sweety439 on 2021-10-19 at 12:47
 2021-10-23, 20:29 #197 sweety439     "99(4^34019)99 palind" Nov 2016 (P^81993)SZ base 36 1100001011012 Posts Base b=72 has largest known minimal prime (start with b+1) for all bases 2<=b<=1024: 3(71^1119849), its value is 4*72^1119849-1, it has 2079933 digits when written in decimal (this number is proven prime, for unproven probable prime, the largest known minimal prime (start with b+1) for all bases 2<=b<=1024 is base b=23, the PRP 9(14^800873), its value is (106*23^800873-7)/11, it has 1090573 digits when written in decimal)
2021-10-23, 20:39   #198
sweety439

"99(4^34019)99 palind"
Nov 2016
(P^81993)SZ base 36

60558 Posts

Quote:
 Originally Posted by sweety439 New minimal prime (start with b+1) in base b is found for b=650: 3:{649}^(498101), see https://mersenneforum.org/showpost.p...&postcount=931 Added it to excel file https://docs.google.com/spreadsheets...RwmKME/pubhtml Base 108 is an interesting base since .... * For the family {1}, length 2 is prime, but the next prime is large (length 449) * For the family 1{0}1, length 2 is prime, the next prime is not known * For the family y{z}, first prime is large (length 411) * For the family 11{0}1, first prime is large (length 400) * For the family {y}z, first prime is large (length 492) (note that length 1 is also prime, but length 1 is not allowed in this project) * For the family 6{0}1, first prime is large (length 16318) * For the family #{z} (# = (base/2)-1)), first prime is large (length 7638) This situation is not common in bases with many divisors, but although 108 has many divisors, this situation occurs in this base, this is why this base is interesting :)) Also base 282 .... * For the family A{0}1, first prime is large (length 1474) * For the family C{0}1, first prime is large (length 2957) * For the family z{0}1, first prime is large (length 277) http://www.noprimeleftbehind.net/cru...82-reserve.htm only tells you that all these three families have a prime with length <= 100001 .... * For the family 7{z}, first prime is large (length 21413) * For the family 10{z}, first prime is large (length 780) http://www.noprimeleftbehind.net/cru...82-reserve.htm only tells you that the farmer family has a prime with length <= 100001, and the letter family has a prime with length <= 100002
Another interesting base is b=23, this base is SOLVED when single-digit primes (i.e. the primes p<23) and the prime "10" (i.e. the prime p=23) are included, and this base is the solved base with the largest minimal prime ((106*23^800873-7)/11, or 9(E^800873), which has 1090573 digits when written in decimal) and the largest number of minimal primes (6021 primes or PRPs, while the second-largest number (base 42) has 4551 proven primes), this base seems be easy and high weight, but in fact this base is low-weight base and this base has large primes for special forms listed in https://docs.google.com/spreadsheets...RwmKME/pubhtml ....

* 1{0}2 (b^n+2): 12 digits [a record value for bases b] (minimal prime in my project, but not minimal prime in original project (i.e. the primes p<=base are also included), since this prime has "10" and "2" as subsequences)
* 4{0}1 (4*b^n+1): 343 digits [a record value for bases b]
* {4}1: 13 digits [not a record value for bases b, since base b=11 requires 45 digits]
* 8{0}1 (8*b^n+1): 119216 digits [a record value for bases b]
* {K}L (extended Sierpinski problem with k>base (k=10)): 3762 digits [a record value for base b for extended Sierpinski problem with k=10 (the prime (10*23^3762+1)/11, exponent is 3762), the previous record is b=17, the prime 10*17^1356+1, exponent is 1356]
* y{z} ((b-1)*b^n-1, Williams prime of the 1st kind): 56 digits [a record value for bases b]
* z{0}1 ((b-1)*b^n+1, Williams prime of the 2nd kind): 15 digits [not a record value for bases b, since base b=19 requires 30 digits, also the same as base b=20, both require 15 digits]
* {z}1 (b^n-(b-1), dual Williams prime of the 1st kind): 17 digits [a record value for bases b, the same as base b=20, both require 17 digits]
* {z}y (b^n-2): 24 digits [a record value for bases b]

Last fiddled with by sweety439 on 2021-10-26 at 22:08

 Similar Threads Thread Thread Starter Forum Replies Last Post sweety439 sweety439 136 2021-11-28 19:42 sweety439 Miscellaneous Math 6 2019-11-25 07:37 davar55 Puzzles 13 2018-03-15 14:46 Flatlander Puzzles 40 2011-02-10 09:42 davar55 Puzzles 5 2008-11-02 00:08

All times are UTC. The time now is 05:04.

Wed Dec 8 05:04:55 UTC 2021 up 137 days, 23:33, 1 user, load averages: 1.94, 1.89, 1.69