Expression syntax
The search box takes far more than a plain number. You can type an arithmetic
expression and the server evaluates it to a single integer, then looks that number up
(or, with Create, adds it). Whitespace is ignored, and multiplication may be left
implicit: 2(3+4) means 2*(3+4). Reading never stores anything.
Every result is capped at about 10 000 000 digits, and each function below has a fixed input ceiling chosen so a single evaluation stays near 3 seconds. Over the ceiling the request is declined rather than run.
Operators
| Operator | Meaning | Example |
|---|---|---|
| + - | Addition, subtraction | 2^100-2^50+1 |
| * / | Multiplication, exact division (the result must be a whole number) | (2^101+1)/3 |
| ^ | Power (right-associative: 2^3^2 = 2^(3^2)) | 2^127-1 |
| % | Modulo (remainder) | 2^100%1000000007 |
| \ | Integer division: a\b = ⌊a/b⌋, the quotient rounded down (same precedence as * / %) | 2^100\7+1 |
| | | Numerator: a|b = a/gcd(a,b), the numerator of the fraction a/b in lowest terms (10|4 = 5, as 10/4 = 5/2; same precedence as * / %) | 100!|10^50 |
| ( ) | Grouping | (10^500-1)/9 |
Intermediate values may be negative, but the final result may not - 3-5 is
rejected. There is no unary minus, with one exception: a function argument may start with a
minus sign, as in U(20,1,-1) (read as 0-1).
Shortcuts
| Form | Meaning | Example |
|---|---|---|
| Mn Fn | Short for M(n) and F(n), the Mersenne and Fermat numbers (see Named functions) | M127 |
| n! | Factorial | 1000! |
| n# | Primorial - product of all primes ≤ n | 100# |
| n## n~ | Product of the first n primes | 10## |
| odd(n) | Odd part of n - n with all its factors 2 removed (A000265) | odd(100!) |
| n? | Least common multiple of 1, 2, …, n - every prime ≤ n to its highest power ≤ n (A003418) | 1000? |
| gcd(a,b,…) lcm(a,b,…) | Greatest common divisor and least common multiple of 2 to 16 arguments; the arguments of one expression may have at most 1,000,000 digits together | gcd(2^600-1,2^900-1) |
| I(n) fib(n) | Fibonacci number | I(5000) |
| Ln lucas(n) | Lucas number | lucas(5000) |
| b,n+ b,n- | Shorthand for bn±1 | 2,127- |
| #id | The stored number with that database id (used inside a larger expression) | #1100000000000000001/31 |
Named functions
Each is written NAME(argument, …). Every argument may itself be an expression,
so Q(5+5) and 2^Q(50)-1 are fine, and functions nest:
Q(Z(10)). The right-hand column is the largest input accepted (matched to the
~3 second budget); for the families marked † a very large parameter is also
declined early when the result would exceed the digit limit.
| Function | Meaning | OEIS | Example | Max input |
|---|---|---|---|---|
| A(n) | Alternating factorial, n!-(n-1)!+… | A005165 | A(100) | n ≤ 2 M |
| B(n) | Bell numbers ‡ | A000110 | B(100) | n ≤ 100 k |
| C(m,n) | Cyclotomic polynomial Φm evaluated at n (C(12,10) = 9901; Φ0 = 1, as 0^0 and 0!) | A019320 A253240 | C(1500,10) | nm ≤ 107 digits |
| D(n) | Partitions into distinct parts ‡ | A000009 | D(1000) | n ≤ 6·107 |
| E(n) | Euler zigzag numbers ‡ | A000111 | E(100) | n ≤ 100 k |
| F(n) | Fermat number 22n+1 | A000215 | F(5) | n ≤ 24 |
| G(m,n) | Numerator of 1/p1m + 1/p2m + … + 1/pnm over the first n primes | A024451/A061015 | G(1,100) | result ≤ 7·106 digits |
| H(m,n) | Metallic m-Fibonacci: F=m·F-1+F-2 (m=1 Fibonacci, m=2 Pell) | A000045/A000129 | H(2,100) | † n ≤ 50 M |
| I(m,n) | m-step Fibonacci (sum of the previous m terms, starting 0,…,0,1; m=2 Fibonacci, m=3 tribonacci) | A000073… | I(3,100) | m3·n ≤ 5.4·108 |
| J(n) | Coefficient of qn of the modular invariant j = 1/q + 744 + 196884q + … ‡ | A000521 | J(100) | n ≤ 107 |
| K(n) | Sum of factorials, 1!+2!+…+n! | A007489 | K(100) | n ≤ 2 M |
| L(m,n) | Metallic m-Lucas: L=m·L-1+L-2, L0=2, L1=m (m=1 Lucas, m=2 Pell-Lucas) | A000032/A002203 | L(2,100) | † n ≤ 50 M |
| M(n) | Mersenne number 2n-1 | A000225 | M(127) | n ≤ 33 M |
| N(m,n) | Multifactorial n·(n-m)·(n-2m)·… (m=1 is the factorial, m=2 the double factorial) | - | N(2,7) | n ≤ 4 M |
| O(n) | Ordered Bell (Fubini) numbers ‡ | A000670 | O(100) | n ≤ 100 k |
| P(n) | Partition numbers § | A000041 | P(1000) | n ≤ 1010 |
| Q(n) | Perrin number | A001608 | Q(1000) | n ≤ 30 M |
| R(m,n) | Repunit - n ones in base m, i.e. (mn-1)/(m-1) | A002275 | R(10,5) | † n ≤ 20 M |
| S(m,n) | Smarandache - concatenate 1,2,…,n written in base m | A007908 | S(10,6) | n ≤ 2.5 M |
| U(n,p,q) | Lucas sequence Un(p,q): U=p·U-1-q·U-2, U0=0, U1=1 | - | U(20,1,-1) | † n ≤ 50 M |
| V(n,p,q) | Lucas sequence Vn(p,q): V=p·V-1-q·V-2, V0=2, V1=p | - | V(20,1,-1) | † n ≤ 50 M |
| W(m,n) | Numerator of 1 + 1/2m + … + 1/nm (m=1 harmonic numbers, m=2 Wolstenholme numbers) | A001008/A007406 | W(2,100) | m·log10(n!) ≤ 5·106 |
| X(m,n) | Smarandache-Wellin - concatenate the first n primes written in base m | A019518 | X(10,100) | † n ≤ 2 M |
| Y(m,n) | m-step Lucas (sum of the previous m terms) | A001644… | Y(3,100) | m3·n ≤ 5.4·108 |
| Z(n) | Motzkin number ‡ | A001006 | Z(500) | n < 20.8 M |
| @(n) | The n-th prime (1-based: @(1) is 2) ‡ | A000040 | @(1000000) | n ≤ 6.5·1012 |
| $(n) | Padovan number | A000931 | $(1000) | n ≤ 40 M |
| &(n) | Narayana's cows | A000930 | &(1000) | n ≤ 30 M |
The Lucas U/V parameters p and q may be negative - U(20,1,-1) is
the Fibonacci numbers (p=1, q=-1) and V(20,1,-1) the Lucas numbers.
With q = 1 or q = -1 the index may be negative too: U(-n,p,q)
= -q-n·U(n,p,q) and
V(-n,p,q) = q-n·V(n,p,q),
e.g. U(-7,3,-1).
Repunit and Smarandache bases must be at least 2, and the multifactorial step at least 1.
I(n) and lucas(n) with one argument stay the
Fibonacci and Lucas numbers; with two arguments they are the m-step Fibonacci and
m-Lucas numbers.
‡ Read from precomputed tables - these values are not computed on request. A value the
table does not hold yet is reported as such. @(n) starts from a table of every
10,000th prime and sieves only the short rest. Z(n) likewise continues from a
table of every 10,000th Motzkin number (with the one before it), at most 9,999 steps.
§ Computed on request (FLINT's Hardy-Ramanujan-Rademacher formula: P(1010), 111,391 digits, takes about 0.4 s). One expression may use at most 4 different n above 109.
Arithmetic functions
These are computed from the complete factorization of their argument n ≥ 1. A number below 264 is factored on the spot; a larger one must be stored here and fully factored (every factor prime or probable prime) - otherwise the expression is not accepted. A number stored with such a term keeps it, and its value is read from that factorization.
| Function | Meaning | OEIS | Example |
|---|---|---|---|
| rad(n) | Radical - the product of the distinct prime factors | A007947 | rad(10!) |
| core(n) | Squarefree part - n/core(n) is a square | A007913 | core(10!) |
| eulerphi(n) | Euler's totient - the numbers up to n coprime to n | A000010 | eulerphi(2^128+1) |
| dedekindpsi(n) | Dedekind's psi - n times the product of (1+1/p) over its primes | A001615 | dedekindpsi(2^128+1) |
| carmichaellambda(n) | Carmichael's lambda - the least m with am = 1 (mod n) for every a coprime to n | A002322 | carmichaellambda(2^256+1) |
| sigma(n,k) | Sum of the k-th powers of the divisors; sigma(n) = sigma(n,1) is the sum of the divisors, sigma(n,0) their number | A000203/A000005 | sigma(2^256+1) |
| omega(n) | Number of distinct prime factors | A001221 | 2^omega(10!)+1 |
| bigomega(n) | Number of prime factors counted with multiplicity | A001222 | 2^bigomega(10!)-1 |