We can modify the idea of the proof by mathematical induction, by first proving P(0) and P(1) then continuing as normal. We can also do, P(0) then assume true for 1, 2, …, n to then prove n + 1
The formula is not defined for n = 1 as $\frac{1}{(n-1) \times n}$ will be $\frac{1}{0 \times 1}$ which is undefined. Because the iteration is not defined, (when we hav 1…0 do we do just 1 or just 0
If we are to do induction, we have to start at n = 2, but. $\frac{1}{1 \times 2} \stackrel{!}{=} \frac{3}{2} - \frac{1}{2}$ $\text{ L.H.S } = \frac{1}{2}$ but R.H.S = 1 CONTRADICTION.
W.T.S ϕn − 2 ≤ Fn where Fn is the nth term in the fibonnaci sequence.
For n = 1, F1 = 1 and $\phi^{1-2} = \frac{2}{1 + \sqrt{5} }$ We know that the square root function is monotonic for numbers greater than 1, and $\sqrt{4} = 2$,so $1+ \sqrt{5} > 1 + 2 = 3$. ⟹ ϕ−1 < 1
For n = 2, F2 = 1 and ϕ0 = 1 and so we havee ϕ0 ≤ F2
For n = 3, F3 = 2 and $\phi^{1} = \frac{1 + \sqrt{5} }{2} \approx 1.618$ and so we have ϕ1 ≤ F3
For n = 4, F4 = 3 and ϕ2 ≈ 2.618 and so we have ϕ2 ≤ F4
With foresight, the inductive step uses 2n − 4 so I wanted to make sure that this doesn’t cause any issues and figured out 1 ≤ n ≤ 4 explicitly.
Fn = Fn − 1 + Fn − 2
As we are dealing with all positive numbers, we have
⟹ ϕn − 3 + ϕn − 4 ≤ Fn.
We need to show ϕn − 2 ≤ ϕn − 3 + ϕn − 4.
Plug in the formula values. $$\text{ LHS } = \frac{\left( 1 + \sqrt{5} \right)^{n-2} }{2^{n-2}}$$
$$\text{ RHS } = \frac{\left( 1 + \sqrt{5} \right)^{n-3} }{2^{n-3}} + \frac{\left( 1 + \sqrt{5} \right)^{n-4} }{2^{n-4}}$$
$$= \frac{\left( 1 + \sqrt{5} \right)^{n-3} }{2^{n-3}} + \frac{2 \times \left( 1 + \sqrt{5} \right)^{n-4} }{2^{n-3}}$$
$$= \frac{\left( 1 + \sqrt{5} \right)^{n-3} + 2 \times \left( 1 + \sqrt{5} \right)^{n-4} }{2^{n-3}}$$
$$= \frac{2 \times \left( 1 + \sqrt{5} \right)^{n-3} + 4 \times \left( 1 + \sqrt{5} \right)^{n-4} }{2^{n-2}}$$
We only need to compare numerators, we WTS, $\left( 1 + \sqrt{5} \right)^{n-2} \le 2 \times \left( 1 + \sqrt{5} \right)^{n-3} + 4 \times \left( 1 + \sqrt{5} \right)^{n-4}$
$u `:=` 1 + \sqrt{5}$
un − 2 ≤ 2un − 3 + 4un − 4 ⟹ u2un − 4 ≤ 2un − 4u + 4un − 4 ⟹ u2 ≤ 2u + 4 We can now plug in the values.
$\implies (1 + sqrt{5})^2 = 6 + 2\sqrt{5} \le 2(1 + sqrt{5}) + 4 = 6 + 2\sqrt{5} \square$
n = 2 we are done as 2 is a prime.
For sake of contradiction, suppose that n cannot be expressed as a product of primes. Why would this be the case? n must have at least 2 factors, as it is greater than 1, and n itself is a factor of n. But it must have more factors, as if it only had 2 factors than n is prime and we are done, we have a expression of a product of primes.
If it had a factor ai = n, i > 1, We must have that a < n, and therefore by our assumption a can be expressed as a product of primes. so n = a = p1…pn. In fact, any factor has the property of being less than n and therefore by our induction step we know it is expressed as a factor of primes. ▫
Wts if a′m + b′n = c, am + bn = d hold before step 4 on the euclidean_algorithm, then they hold after.
Set c ← d,
We have d = a′m + b′n
d ← r, We have am + bn = r
t ← a′, a′ ← a, We have d = am + b′n
a ← t − qa,
t ← b′,
b′ ← b,
b ← t − qb
Go to step 2.
We have d = a′m + b′n
| t | a | a’ | b | b’ | m | n | c | d |
|---|---|---|---|---|---|---|---|---|
| a’ | t-qa | a | d | r | ||||
| b’ | t-qb | b |
am + bn = d and am + bn = r
F1 = 12 = 1 F2 = 22 − 12 = 3 = F1 + 2 F3 = 32 − 22 + 12 = 6 = F2 + 3 F4 = 42 − 32 + 22 − 12 = 10 = F3 + 4 F5 = 52 − 42 + 32 − 22 + 12 = 15 = F4 + 5
Fk + 1 = (k + 1)2 − Fk = k2 + 2k + 1 − Fk Fk = Fk − 1 + k tautologically ⟹ Fk + 1 = k2 + 2k + 1 − Fk − 1 − k Fk = k2 − Fk − 1 ⟹ Fk + 1 = Fk + k + 1
13 = 1 23 = 3 + 5 33 = 7 + 9 + 11 43 = 13 + 15 + 17 + 19 etc.
f(k) = k * (k − 1) + 1 + k * (k − 1) + 3 + k * (k − 1) + 4
(k + 1)2(k) + k + 1 + k(2k + 1)
$f(k+1) = $ = k3 + 3k2 + 3k + 1
RHS (k + 2)3 = (k2 + 4k + 4)(k + 2) = k3 + 4k2 + 4k + 2k2 + 8k + 8 k3 + 6k2 + 12k + 8
LHS (k2 + 2k + 1)(k + 1) + 3(k2 + 2k + 1) + 3k + 3 + 1 = k3 + 2k2 + k + k2 + 2k + 1 + 3k2 + 6k + 3 + 3k + 3 + 1 = k3 + 6k2 + 12k + 8
13 = 12 13 + 23 = 9 = (1 + 2)2 13 + 23 + 33 = 36 = (1 + 2 + 3)2
We assume the that the formula $\sum _{i=1}^{n} i^3 = \left( \sum _{i=1}^{n} i \right)^2$ holds true.
$$\sum _{i=1}^{n} i^3 = \sum _{i=0}^{n-1}i^3 + 3i^2 + 3i + 1$$ $$= (\sum _{i=1}^{n-1} i)^2 + 3 \sum _{i=1}^{n-1} i^2 + 3 \sum _{i=1}^{n-1} i + n$$ $$\left( \frac{1}{2}(n)(n-1) \right) ^2 + 3 \frac{1}{6}(n-1)(n)(2(n-1)+1) + 3 \frac{1}{2}(n-1)(n) + n$$ $$= \frac{1}{4}n^2(n^2 - 2n + 1) + \frac{1}{2}(n^2 - n)( 2n -1) + \frac{3}{2}(n^2 -n) + n$$ $$\frac{1}{4} (n^{4} - 2n^3 + n^2) + \frac{1}{2}(2n^3 - n^2 - 2n^2 + n) + \frac{3}{2}(n^2 - n) + n$$ $$\frac{1}{4} (n^{4} - 2n^3 + n^2 + 2(2n^3 - n^2 - 2n^2 + n) + 6(n^2 - n) + 4n)$$ $$\frac{1}{4} (n^{4} - 2n^3 + n^2 + 4n^3 - 2n^2 - 4n^2 + 2n + 6n^2 - 6n + 4n)$$ $$\frac{1}{4} (n^{4} + 2n^3 + n^2)$$
Compare to $$\left( \frac{1}{4}a^2(a+1)^2 \right)$$ $$\frac{1}{4}a^2 (a^2 + 2a + 1)$$ $$\frac{1}{4} (a^{4} + 2a^3 + a^2)$$
Which is in the form we wanted ▫
We have to prove this for integers positive and negative. Let’s start with the positive integers.
n = 0 ⟹ 1 ≥ 1 n = 1 ⟹ (1 − a) ≥ 1 − a n = 2 ⟹ (1 − a)2 = 1 − 2a + a2 ≥ 1 − 2a
Assume that for n = k ⟹ (1 − a)k ≥ 1 − ka to show true for n = k + 1 WTS. (1 − a)k + 1 ≥ 1 − (k + 1)a LHS. (1 − a)k + 1 = (1 − a)k(1 − a) RHS. 1 − (k + 1)a = 1 − ka − a
By our inductive step, we know that (1 − a)k ≥ (1 − ka) so we WTS, $$(1-a)^{k}(1-a) \ge (1-ka)(1 - \frac{a}{1-ka})$$ We cancel out the known terms $$\iff(1-a) \ge (1-\frac{a}{1-ka})$$ $$\iff -a \ge -\frac{a}{1-ka}$$ $$\iff 1 \le \frac{1}{1-ka}$$ ⇔ 1 − ka ≤ 1 ⇔ −ka ≤ 0 which is part of our assumption of a hence true.
Inductive step. Assume true that for n = −k ⟹ (1 − a)−k ≥ 1 + ka to show true for n = −k − 1
WTS. $$\frac{1}{(1-a)^{k}} \times \frac{1}{1-a} \ge!! 1 -(-k-1)a = 1 + ka + a$$
RHS $$1 + ka + a = (1+ka) (1+ \frac{a}{1+ka})$$
WTS $$\frac{1}{1-a} \ge 1 + \frac{a}{1 +ka}$$ $$\iff \frac{1}{1-a} \ge \frac{1 + ka + a}{1 +ka}$$ ⇔ 1 + ka ≥ (1 + ka + a)(1 − a) = 1 + ka − ka2 − a2 ⇔ 0 ≥ −ka2 − a2 which is true.
TODO readable and potentially wrong in proof We need to look up nature of cubic equations.
if n = 10 then 210 = 1024 and n3 = 1000.
2k + 1 = 2k21. We know 2k > k3 so this implies 2k21 > 2k3
Suffice to show 2(k)3 > (k + 1)3. (k + 1)3 = (k + 1)(k2 + 2k + 1) = (k3 + 2k2 + k + k2 + 2k + 1) = k3 + 3k2 + 3k + 1
2k3 − k3 − 3k2 − 3k − 1 $f(k) `:=` k^3 - 3k^2 - 3k - 1$
$\frac{d}{d k} f = 3k^2 - 6k - 3$ wts that the function is increasing when k ≥ 10
consider $g(k) `:=` k^2 - 2k - 1$
$\frac{d}{d k} g(k) = 2k - 2$ which is increasing for k > 10
⟹ g(k) is increasing, ⟹ f(k) is increasing, hence the our original statement is must be true, ▫