There Are 42 Ways to Break Ten Into Parts, and 190 Million for a Hundred
Published 10/26/2026 · 3 min read · Everyday calculators
Daniel Okonkwo — Front-end developer and tech writer at OneKitly
Web performance · File formats
Checked against 2 sources
A partition of a number is a way of writing it as a sum of positive whole numbers where the order does not matter, so four can be written as 4, 3+1, 2+2, 2+1+1, and 1+1+1+1 — five ways. For ten there are 42, small enough to list on a page. For fifty there are 204,226. For a hundred there are 190,569,292, and for two hundred just under four trillion. The growth is the point: doubling the input from a hundred to two hundred multiplies the answer by about twenty thousand, which is far faster than any polynomial and far slower than doubling per unit. There is no simple closed formula, and the asymptotic expression that describes the growth involves the square root of n in an exponent.
p(10) = 42, p(50) = 204,226, p(100) = 190,569,292, p(200) = 3,972,999,029,388. Doubling the input from a hundred to two hundred multiplies the count by twenty thousand.
An approximation that is right to within five per cent
Hardy and Ramanujan produced an asymptotic formula in 1918 whose leading term already tracks the truth closely: it gives about 199 million where the exact answer is 190,569,292, an overestimate of 4.6 per cent, and about 4.10 trillion against the exact 3.97 trillion at two hundred, an overestimate of 3.2 per cent. The error shrinks as n grows, which is what asymptotic means, and it never vanishes. Rademacher later turned the approximation into an exact convergent series, so the value can be computed to any precision — but the short formula remains the one worth remembering, because it makes the shape of the growth visible in a way an exact series does not.
Order not mattering is what makes it hard
If order did matter the answer would be trivial: the number of ordered sums of positive integers making n is exactly two to the power n−1, a formula anyone can derive in a line. Insisting that 3+1 and 1+3 are the same partition destroys that structure and leaves a counting problem with no product formula, no simple recursion on n alone, and no closed form. It is a recurring lesson in combinatorics that removing a distinction usually makes counting harder rather than easier, because the objects being counted lose the symmetry that a formula would have exploited.
| n | p(n) | Digits |
|---|---|---|
| 10 | 42 | 2 |
| 50 | 204,226 | 6 |
| 100 | 190,569,292 | 9 |
| 200 | 3,972,999,029,388 | 13 |
Frequently asked questions
- How is the exact value computed, then?
- Usually by a recurrence that builds every value from the ones before it, using a formula whose terms alternate in sign and whose indices come from a sequence of pentagonal numbers. It is fast enough that a laptop reaches n in the thousands without effort, and it is the reason a tool can return p(500) instantly without any clever mathematics at run time. The asymptotic formula is for understanding and the recurrence is for computing, which is a division of labour that shows up all over numerical work.
- Do partitions come up outside pure mathematics?
- Yes, wherever indistinguishable things are grouped and the grouping is what matters. The number of distinct energy configurations of identical particles at a given total is a partition count, which puts the function into statistical physics; the number of distinct shapes a set of identical items can take when split across bins is one too, which puts it into scheduling and packing. In each case the difficulty is the same as here — the objects are interchangeable, so the tidy formulas that assume they are labelled do not apply.
Articles you may find interesting
All guides →Related tools
Sources
Spotted a mistake in this article?