Introduction to sequences and series

A sequence is very simple thing to understand. Its basically just a list of numbers – there doesn’t
have to be a pattern or anything of that sort. It can be a random list of numbers, for example
1, 5, 8, 7, 4, 6 can be a sequence by all means.

The key thing note here is that in a sequence the order matters. You can’t just jumble the numbers
around and say the sequence hasn’t changed. For this purpose, we can assign numbers to a
sequence, to make it easier to order the numbers and to call out any number.

• Infinte sequence: 𝑎1 , 𝑎2 , 𝑎3 , 𝑎4 …
• Finite sequence (n terms): 𝑎1 , 𝑎2 , 𝑎3 , 𝑎4 … 𝑎𝑛
What we have done in the above example is that we have named each term in a sequence using a
subscript that shows its position in the list. This ‘ordering’ gives rise to two types of sequences, an
infinite one and a finite one (with n terms).

Just to give you guys a formal mathematical definition about what a sequence is (this definition is
the summary of the previous paragraph):
A sequence is a function whose domain is the set of natural numbers {0,1, 2, 3, 4,…} (or {1, 2, 3, 4, …}).

A sequence can start with the first number being on position 0 or the first number being on position
1. For example, in the above sequence, if we follow the 0 – first method 1 will be the 0th number and
5 will be the 1st number. If we follow the 1 – first method 1 is the 1st number while 5 is the 2nd
number.

Another thing that will be useful to our investigations is a partial sum sequence. The partial sum of a
part of a sequence. By summing up the terms of a sequence we can form a new sequence. For
example, Let’s say the base sequence is 𝑆𝐴 and the partial sum sequence is 𝑆𝐵. The n term of the 𝑆𝐵
sequence will equal:

SBn = SA1 + SA2 + SA3 + SA4 + ⋯ + SAn

The partial sum sequence of the sequence 4, 6, 8, 13, 15 will be 4, 10, 18, 31, 46.

Some terminology

Recursive formula: A recursive formula is a formula that gives you the 𝑛𝑡ℎ term of a sequence.
However, a recursive formula uses the previous term for this purpose. If you want to find the 2nd
term you will have to know the 1st term. In general, if you want to find the 𝑛𝑡ℎ term of a
sequence using a recursive formula you will have to consider the (𝑛 − 1)𝑡ℎ term.

We will see that finding this formula will be very easy – we just need a relation between any
consecutive numbers in a sequence. Though it is easy to find it is not a really appreciable formula
– after all if you want to say the 1000th term you need to know the 999th term of a sequence.

Closed formula: A closed formula is the ultimate formula. A closed is the way you would think
about a formula – you don’t need to input the previous number; you only have to input what
term you want and some other easy-to-get information – and TA-DA the formula spits out the
term you want. This formula is a little tricky to derive.
 Types of Sequences
I have broadly studied three types of sequences: Arithmetic sequences (broadly k-constant
sequences, we’ll get into that later), geometric sequences and Fibonacci-type sequences. Let’s study
them in great detail.

Arithmetic Sequences

An arithmetic sequence is a sequence that has a common difference. A difference is equal to the
difference of any two consecutive terms in an arithmetic sequence. In an arithmetic sequence this
difference is the same in any case. For example, take the sequence 8, 13, 18, 23, 28. You can see that
the difference between any two consecutive terms is 5:

8 (+𝟓) 13 (+𝟓) 18 (+𝟓) 23 (+𝟓) 28

Using this above pattern, we can easily come up with a general recursive formula for an arithmetic
sequence. We know that difference between any two consecutive terms is constant. That means
that if you want the 𝑛𝑡ℎ term we have to add the common difference, 𝑑 to the previous term, 𝑛 − 1.
We can put this mathematically as:

𝒂𝒏 = 𝒂𝒏−𝟏 + 𝒅
We could also have used some simple algebra for this problem. Since we know that: 𝑎𝑛 − 𝑎𝑛−1 = 𝑑
we can transpose −𝑎𝑛−1 to the right-hand side to get: 𝑎𝑛 = 𝑑 + 𝑎𝑛−1 .

Let’s find out the 5th term in the sequence using the recursive formula: 8, 13, 18, 23 … Here n will be
5, d will be 5 and the previous number will be 23:

• 𝐒𝐭𝐞𝐩 𝟏: 𝑎5 = 𝑑 + 𝑎5−1
• 𝐒𝐭𝐞𝐩 𝟐: 𝑎5 = 𝑑 + 𝑎4
• 𝑺𝒕𝒆𝒑 𝟑: 𝑎5 = 23 + 5
• 𝐓𝐡𝐞 𝟓𝐭𝐡 𝐭𝐞𝐫𝐦 𝐢𝐬 𝟐𝟖
Now let’s try to find a closed formula for an arithmetic sequence. For finding a formula we need to a
spot a pattern – then we generalize it.

Hmmm… what if we try to find the relationship between the position of a term in a sequence with
the number of times we add the common difference to get that term?

It’s not an idea pulled out of the air – the difference between two consecutive terms is equal to the
difference, the difference between alternate terms will be twice the difference, and the difference
between the nth term and the n+3 term will equal thrice the difference - so there has to be some
relationship with the term number and the number of times you add the difference.

Let’s use the above sequence and see how many times you have to add the difference, 5, to get the
nth term of the sequence: 8, 13, 18, 23, 28 To help us spot the patterns fast let’s draw a table with
our results.

Term position Term Break up

1 8 8+5∗0
2 13 8+5∗1
3 18 8+5∗2
4 23 8+5∗3
5 28 8+5∗4
For the first number which is 8, you don’t add 5 at all – you just add the starting number 8! For the
second number you have to add 5 and 8. Going to the third number you have to add 5 twice and 8 to
that.

Pay attention to the break up. You add 8 always – so that becomes part of our formula. Now let us
have a look at how many times we add five. If you see the multiplier of 5 is exactly one less than the
position of the term in each case! So, if you want the nth term in sequence the formula becomes:

8 + (𝑛 − 1)5
Now we can take this formula one step ahead – lets try to form a general closed formula for the nth
term of a AS (arithmetic sequence). We can replace 8 with 𝑎, which is the first term of a AS. 5 is just
the common difference of this sequence so we can replace it with another variable say 𝑑. The
formula becomes:

𝒂 + (𝒏 − 𝟏)𝒅
Isn’t this cool! We have a general closed formula for absolutely any arithmetic sequence! All we
need to know is the first term and common difference and we can get any term of the sequence. For
fun, let’s try this formula on a random sequence, say: 11, 2, −7, −16, −25 and we want the 76th
term of this sequence. Here 𝑎 = 11, 𝑑 = −9 𝑎𝑛𝑑 𝑛 = 76:

• a + (n − 1)d
• 8 + (76 − 1) − 9
• 8 + 75 ∗ −9
• 8 − 675
𝐓𝐡𝐞 𝟕𝟔𝐭𝐡 𝐭𝐞𝐫𝐦 𝐨𝐟 𝐭𝐡𝐞 𝐬𝐞𝐪𝐮𝐞𝐧𝐜𝐞 𝐢𝐬 − 𝟔𝟔𝟕
(Isn’t this much better than just memorizing some formula?)

What this formula can’t do

This above formula is pretty general – but it does have certain limitations.

This formula ONLY works if the differences between consecutive numbers is

constant (you may have realized this as this is how we designed the formula)

For example, let’s say we wanted to figure out the nth term of the partial sum
sequence of the arithmetic sequence: 1, 2, 3, 4. The partial sum sequence will be the
triangular numbers: 1, 3, 6, 10.

If we wanted to find the nth term of the triangular numbers we couldn’t use the
formula we just derived. This is because the differences are not constant – the
differences are the natural numbers (because we added them up).

1 (+2) 3 (+3) 6 (+4) 10

So, is there is a formula for the sum of a arithmetic sequence? Stay tuned to find
out.
 Sum of an arithmetic sequence
Now I want to find a formula for the nth term of a partial sum sequence of an arithmetic sequence.
How do we get started? By spotting a pattern.

The pattern that I stumbled upon that helped me find the formula was related to the sum of the
‘pairs’ in an arithmetic sequence. Let me explain in a bit more detail. Consider our old arithmetic
sequence:

𝟖, 𝟏𝟑, 𝟏𝟖, 𝟐𝟑
If you see the sum of the red numbers is 31. The sum of the blue numbers is 31 too. In fact, the value
of any pairs in an arithmetic sequence selected this way is equal to the sum of any the pair. Let’s
expand the above sequence to see if it still holds.

𝟖, 𝟏𝟑, 𝟏𝟖, 𝟐𝟑, 𝟐𝟖

The value of the red pair is 36 and the sum of the blue pair is 26 too! So, it does hold! However, 18
does not have any pair so it can’t be summed up. Observe that 18 is the half of 36. Is the middle
number always half of the pair value in a sequence that has an odd number of terms? Let us expand
this sequence further to test out our conjecture.

𝟖, 𝟏𝟑, 𝟏𝟖, 𝟐𝟑, 𝟐𝟖, 𝟑𝟑, 𝟑𝟖

Again, the sum of the red pair = sum of the green pair = sum of the blue pair which equals 46. Ad 46
is exactly twice of 23! So, till now it does seem true that:

• The sum of any pair in an arithmetic sequence equal the sum of any other pair (chosen in the
same way).
• In a sequence that has an odd number of terms the middle term is exactly half of the value
of the sum of any pair.

Now using these statements, we can start making some progress on deriving the formula. Fist of all
we need to figure out what the value of any pair is. Then we need to figure out how many times we
multiply the pair value to get the sum of the sequence.

Finding the sum of any pair: We know that the sum of any pair equals the sum of any pair. So, lets
try to find the sum of the first number of a AS and the last number of AS. (Why am I choosing this
pair? Is it special? No, it is just because it is easy to come up with a formula for the sum of this
pair.)The first number of the AS will be 𝑎 the last number will be what? Well can’t we just use the
formula for the nth term of a AS directly? Yes we can! This is because the formula gives you the nth
term of an arithmetic sequence and that’s what we are looking for. So, we add he two to get:

𝑎 + 𝑎 + (𝑛 − 1)𝑑 = 2𝑎 + (𝑛 − 1)𝑑
Now we need to figure out how many times we multiply this value. For sequence that has an even
number of terms this is easy. We just multiply this equation by half the number of terms or half of n.
𝑛
It turns out this is true for odd numbers too – we can multiply the sum for pair value by 2 . This is
because the middle term in a sequence which has an odd number of terms is half. For example, if we
have a AS whish has three terms, you will have a pair that equals: 2𝑎 + (𝑛 − 1)𝑑 and term that
2𝑎+(𝑛−1)𝑑 3
equals: . This equal 2𝑎 + (𝑛 − 1)𝑑 ∗ . You can do this for a sequence that ha 5 terms. You
2 2
2𝑎+(𝑛−1)𝑑
will multiply 2𝑎 + (𝑛 − 1)𝑑 by 2 (as there are 2 whole pairs). Then we add . This is the
2
 5
same thing as 2𝑎 + (𝑛 − 1)𝑑 ∗ . So, the Formula for the nth term in the partial sum sequence of a
2
arithmetic sequence:
𝑛 𝑛(2𝑎 + (𝑛 − 1)𝑑)
2𝑎 + (𝑛 − 1)𝑑 ∗ 𝐎𝐑
2 2
Isn’t this fun – formula arise from patterns!

The formula for triangular numbers

If you notice that we can now find a formula for the sum of natural numbers i.e., the
triangular numbers. So first let sum find out the values of variables in the formula:

• 𝑎, the starting number of the AS will be 1, as the natural number sequence

starts with 1 in this case.
• 𝑑, the common difference will be 1, as the natural numbers increase by 1
each time: 1 (+1) 2 (+1) 3 (+1) 4
• 𝑛, the nth term we want will remain n, since we don’t want any number in
the triangular numbers, but we want the nth term.

Substituting and simplifying the formula:

𝑛(2𝑎 + (𝑛 − 1)𝑑) 𝑛(2 ∗ 1 + (𝑛 − 1)1) 𝑛(2 + 𝑛 − 1) 𝒏(𝒏 + 𝟏)
= = =
2 2 2 𝟐
Try the formula out! The formula will now make much more sense since you know
how it is derived – but in maths there are many ways to get to the same answer –
and later in this paper you will see many ways of coming up with the formula.

Dealing with higher partial sums and Polynomial fitting

What if want to want to find a formula for the nth term of the sum of sum of an arithmetic
sequence? Like let’s say we have the arithmetic sequence: 1, 2, 3, 4, 5. And we sum up these
numbers twice. Look at the table below (Sum 2 sequence is the partial sum of Sum one):

Sequence: 1 2 3 4 5 6
Sum one: 1 3 6 10 15 21
Sum two: 1 4 10 20 35 56

Well… the answer to this question is so vast that we have an entire procedure for dealing with these
kinds of partial sums. Before diving into the details lets adopt some formal terminology.

If you see an arithmetic sequence has its first differences constant. A partial sum sequence of an
arithmetic’s sequence has its second differences constant. A partial sum of a partial of a arithmetic
sequence has its second differences constant. So, we say that:

𝚫𝐤 − 𝐜𝐨𝐧𝐬𝐭𝐚𝐧𝐭 if the kth differences are constant

So, sequence one in the above table will be 𝚫𝟐 and sequence two will be 𝚫𝟑 and so on.
Now lets us put this in terms of formula both related to the type of sequences:

• A sequence that is 𝚫𝟎 constant will have a very easy formula-the sequence itself is
constant (like 6,6,6).
• A sequence that is 𝚫𝟏 will be an arithmetic sequence and we have a closed formula for
that.
• A sequence that is 𝚫𝟐 will be the sum of an arithmetic sequence and we have closed
formula for that too.

Now there is a very interesting pattern in the formulas. Allow me to explain:

Every time we increase the value of k, the degree of the formula increases. The formula for 𝛥1
sequences has no powers or anything (look at the formula). That means the degree is 1. The formula
for 𝚫𝟐 is quadratic which means it has squares involved. Look at the formula for the sum of
arithmetic sequences-n times n –which automatically gives rise to quadratic equations. In general,
we can say that:
The closed formula for a sequence for a sequence will be a degree 𝑘 polynomial if and only the sequence is 𝛥𝑘 .

This is again nothing but a pattern!

The above table tells you the functions of the formulas. The alphabets are all coefficients. N is the
nth term. For deriving a formula for a sequence, we have to change the value of n and list out the
expression. For example, consider the triangular numbers (the function will be quadratic):

Now we basically have to solve for a b and c. How do we do it? Well, if we subtract consecutive
expressions c will get eliminated. Only a and b will be left:
Now we can subtract the expression again to get rid of b and we can now find the value of a:

Now we can substitute back and find the value of b:

Since half plus half equals 1, c is actually 0.

Now we can just substitute the values in the above equation. We get:

𝑛2 𝑛 𝑛(𝑛 + 1)
+ =
2 2 2
We can find formula for the sum of the triangular numbers in a similar way (see notes – formula for
tetrahedron numbers). But is there a formula that tells you the nth term kth partial sum of an
arithmetic sequence? Get ready for a long detour.

General formula for Triangular numbers and 𝚫𝟎 sequences

Let’s try to something very general: a formula for the nth term in a kth partial sum of the triangular
numbers.

Step 1: The first step in deriving this formula was trying to see a pattern in the factored formulas for
the sums of the triangular numbers. I derived all formulas till the 3rd sum of the triangular numbers.
See the formulas below:

Formula for sum of triangular numbers:

Formula for sum of triangular numbers (factored):

Formula for sum of the sum of triangular numbers:

Formula for sum of the sum of triangular numbers (factored):

Formula for sum of sum of the sum of triangular numbers:

Formula for sum of sum of the sum of triangular numbers (factored):

It was not easy factoring the formulas – I could factor the formula of the sum of the triangular
numbers using a well-known relationship.
For finding the factored version of the sum of sum of the triangular numbers I dived the formula for
the sum of the sum of the triangular numbers with the formula for the sum of the triangular
numbers. I got 𝑛 + 3. I then multiplied this result with the already factored form of the sum of the
triangular numbers so the numerator become: 𝑛(𝑛 + 1)(𝑛 + 2) ∗ (𝑛 + 3) = 𝑛(𝑛 + 1)(𝑛 + 2)(𝑛 + 3)

I did a similar thing for the next formula. I divided the 3rd formula in the list by the second one two
get: (𝑛 + 4). The whole numerator became: 𝑛(𝑛 + 1)(𝑛 + 2)(𝑛 + 3)(𝑛 + 4)

Step 2: I could see a pattern in the numerator and in the denominator. The numerator was n
multiplied by consecutive numbers and the denominator part was the factorial sequence. But
𝑛(𝑛+1)
something didn’t fit right. The formula for the triangular numbers is but if we consider the
2
triangular numbers are the first sequence 1! Is not 2. Then I realized that the triangular numbers
wasn’t the base sequence, since it could be expressed a sequence of partial sums of the natural
numbers. In fact, even the natural numbers aren’t a base sequence since it can be expressed a
sequence of partial sums! See the list below:

Sequences Formulas

𝐒𝐞𝐪𝐮𝐞𝐧𝐜𝐞 𝟎: 1, 1, 1, 1 … 1
1
𝐒𝐞𝐪𝐮𝐞𝐧𝐜𝐞 𝟏: 1, 2, 3, 4 … 𝑛
1
𝐒𝐞𝐪𝐮𝐞𝐧𝐜𝐞 𝟐: 1, 3, 6, 10 … 𝑛(𝑛 + 1)
2
𝐒𝐞𝐪𝐮𝐞𝐧𝐜𝐞 𝟑: 1, 4, 10, 20, 35 … 𝑛(𝑛 + 1)(𝑛 + 2)
6
𝐒𝐞𝐪𝐮𝐞𝐧𝐜𝐞 𝟒: 1, 5, 15, 35, 70 … 𝑛(𝑛 + 1)(𝑛 + 2)(𝑛 + 3)
24

This ordering made much more sense – Because the factorials fitted up and even the sequences
were labelled in a more meaningful way: Because seq 0 is constant sequence. Seq 1 has its first
differences constant. Sequence 2 had its second differences constant etc. In general

𝐒𝐞𝐪𝐮𝐞𝐧𝐜𝐞 𝐤 𝐢𝐬 𝐚 𝚫𝐤 𝐬𝐞𝐪𝐮𝐞𝐧𝐜𝐞
Step 3: So, the denominator simply becomes: 𝑘! The greatest number part in any formula is
equivalent to 𝑘 − 1. For example, the greatest number added with n in sequence 2 is 1 and the
greatest numbers which is added to n is 2 in sequence 3 etc.

Since the numbers are being multiplied we can take use of the factorial function. We can first do
(𝑛 + 𝑘 − 1)! This is however much greater. Because it starts with n, we have to divide this by
(𝑛+𝑘−1)! 𝑎
(𝑛−1)! 𝑏 𝑎 (𝑛+𝑘−1)!
(𝑛 − 1)! So the formula becomes: . Since = . This is equal to: (𝑛−1)!𝑘!
𝑘! 𝑐 𝑏𝑐

Now we can expand this formula to make it even more general. Since we are really summing up a
constant sequence (with the constant one) k times, we can multiply the whole formula by a, which is
the constant of the first sequence. For example, the sum of the seq 2,2,2,2… will be the even
numbers which is exactly two times the natural numbers. Same goes with 3. The sum of the seq
3,3,3,3,…. Will give you the table of 3 which is thrice the natural numbers. Hence the general
formula becomes:
(𝒏 + 𝒌 − 𝟏)!
𝒂( )
(𝒏 − 𝟏)! 𝒌!

Just for fun I created a formula generator – which gives you a general formula for the kth partial sum
of a constant sequence. Again, the denominator remains the same: k!. Since we want a formula not
the value we can use a pi notation to give us: 𝑛 ∗ 𝑛 + 𝟏 ∗ 𝑛 + 𝟐 ∗ … ∗ 𝑛 + 𝒌 − 𝟏. Here the n value
will remain the same - we need the function to change the value in bold. I used a dummy variable z
for this purpose. So, the bottom value of the pi will be z=0 and the top value will be k-1, with the
formula being: (𝑛 + 𝑧). Compiled:

∏𝒌−𝟏
𝒛=𝟎 (𝒏 + 𝒛)
𝒂
𝒌!
This thing is cool – you can say goodbye to polynomial fitting for finding complex formulas. But these
two formulae only work for 𝚫𝟎 sequences.

Some Examples

Using the formula to figure out the 6th term in the 5th partial sum sequence of the sequence 3,3,3,3…
Here 𝑛 = 6, 𝑘 = 5 and 𝑎 = 3:

(𝒏 + 𝒌 − 𝟏)! (𝟏𝟎)! 𝟑𝟎𝟐𝟒𝟎

𝒂( ) = 𝟑( )=𝟑 = 𝟕𝟓𝟔
(𝒏 − 𝟏)! 𝒌! 𝟓! 𝟓! 𝟏𝟐𝟎 ∗ 𝟏𝟐𝟎

Using the formula generator to come up with the formula for the above example:

∏𝒌−𝟏
𝒛=𝟎 (𝒏 + 𝒛) ∏𝟒𝒛=𝟎(𝒏 + 𝒛) 𝒏(𝒏 + 𝟏)(𝒏 + 𝟐)(𝒏 + 𝟑)(𝒏 + 𝟒)
𝒂 =𝟑 =𝟑
𝒌! 𝟏𝟐𝟎 𝟏𝟐𝟎
Formula for an arithmetic sequence

For deriving a formula for the kth partial sum of an arithmetic sequence, I t would be a good idea to
compare a 𝛥0 with a arithmetic sequence. I compared two sequences with each other. First the
2,2,2,2… seq with the 1, 3, 5, 7… seq. See the table below:

Table #1
 Table #2

Before delving into the relationship between the two sequences, I am going to tweak the previous
formula to give me the Sequence two. I will have to tweak the previous formula as the First partial
sequence is of a Δ0 sequence is the 0 partial sum of an arithmetic sequence. So, we have increase all
the k variables in the previous formula by 1:

(𝒏 + 𝒌 − 𝟏)! (𝒏 + 𝒌)!
𝒂( ) = 𝒅( )
(𝒏 − 𝟏)! 𝒌! (𝒏 − 𝟏)! (𝒌 + 𝟏)!

I have also replaced the constant term by the differences in an arithmetic sequence. That is Part 1 of
the formula one.

Step 2: Now for the relationship. If you notice I have not compared any random sequences. I have
made sequences out of the differences of arithmetic ones. For example, the first arithmetic
sequence is 1, 3, 5, 7 i.e the odd numbers. The constant differences is 2. I have then done the partial
sum of 2,2,2,2.. and compared it. Same with the second example. If you see differences between
(𝒏+𝒌−𝟏)!
the two are equal to the previous formula! More precisely it equals (𝒂 − 𝒅) ( (𝒏−𝟏)!𝒌! )

So now we can compile it as:

(𝒏 + 𝒌)! (𝒏 + 𝒌 − 𝟏)!
𝒅( ) + (𝒂 − 𝒅) ( )
(𝒏 − 𝟏)! (𝒌 + 𝟏)! (𝒏 − 𝟏)! 𝒌!

I created the formula generator too. As the second part of this equation is the same as the previous
one we can just put the same pi notation:

∏𝒌−𝟏
𝒛=𝟎 (𝒏 + 𝒛)
(𝒂 − 𝒅)
𝒌!
(𝑛+𝑘)!
Now for this part 𝑑 ((𝑛−1)!(𝑘+1)!) we can just change the denominator to one more than k and
change the upper k value to k instead of k-1:

∏𝒌𝒛=𝟎(𝒏 + 𝒛) ∏𝒌−𝟏
𝒛=𝟎 (𝒏 + 𝒛)
𝒅 + (𝒂 − 𝒅)
(𝒌 + 𝟏)! 𝒌!
This is the ultimate formula for an arithmetic formula!
 Some Examples

Using the formula to figure out the 6th term in the partial sum sequence of the sequence 1,3,5,7,9…
Here 𝑛 = 6, 𝑘 = 2, 𝑑 = 2 and 𝑎 = 1:
(𝒏 + 𝒌)! (𝒏 + 𝒌 − 𝟏)! (𝟔 + 𝟐)! (𝟔 + 𝟐 − 𝟏)!
𝒅( ) + (𝒂 − 𝒅) ( ) = 𝟐( ) −𝟏( )
(𝒏 − 𝟏)! (𝒌 + 𝟏)! (𝒏 − 𝟏)! 𝒌! 𝟓! 𝟑! 𝟓! 𝟐!

𝟒𝟎𝟑𝟐𝟎 𝟓𝟎𝟒𝟎
𝟐( )− = 𝟏𝟏𝟐 − 𝟐𝟏 = 𝟗𝟏
𝟕𝟐𝟎 𝟐𝟒𝟎
Using the formula generator to come up with the formula for the above example:

∏𝒌𝒛=𝟎(𝒏 + 𝒛) ∏𝒌−𝟏
𝒛=𝟎 (𝒏 + 𝒛)
𝒅 + (𝒂 − 𝒅)
(𝒌 + 𝟏)! 𝒌!
∏𝟐𝒛=𝟎(𝒏 + 𝒛) ∏𝟏𝒛=𝟎(𝒏 + 𝒛) 𝒏(𝒏 + 𝟏)(𝒏 + 𝟐) 𝒏(𝒏 + 𝟏)
𝟐 + − =𝟐 + −
𝟔 𝟐 𝟔 𝟐
𝟒𝒏 + 𝟔𝒏𝟐 + 𝟐𝒏𝟑 −𝒏 − 𝒏𝟐 𝟒𝒏 + 𝟔𝒏𝟐 + 𝟐𝒏𝟑 − 𝟑𝒏 − 𝟑𝒏𝟐 𝒏 + 𝟑𝒏𝟐 + 𝟐𝒏𝟑
+ = =
𝟔 𝟐 𝟔 𝟔
Interesting observations in triangular numbers

There are numerous ways to come up with a closed formula for the sum of natural numbers i.e
triangular numbers. We could use the sum of an arithmetic sequence formula since it is the sum of
natural numbers which is an arithmetic sequence. Using the formula:

𝐧(𝟐𝒂 + 𝒅(𝒏 − 𝟏))

𝟐
The other way is to think about the formula for the area of a triangle. The formula is half times the
(𝑏𝑎𝑠𝑒)(ℎ𝑒𝑖𝑔ℎ𝑡)
base times the height or 2
. If you observe figure 1 carefully you will notice that the
length of any triangle is n and (n+1). We can substitute these values into the above formula to get:
𝑛(𝑛 + 1)
2
An interesting property of triangular numbers is that when you sum 𝑛 and 𝑛 + 1 numbered
triangular numbers you get (𝑛 + 1)2 . This can prove algebraically with the help of the formula for
the T numbers:
𝑛(𝑛 + 1) (𝑛 + 1)(𝑛 + 2) (𝑛 + 1)(2𝑛 + 2)
+ = = 𝑛2 + 2𝑛 + 1
2 2 2
We get a peculiar answer, which on first sight doesn’t prove the above relationship, but let us go
step by step and decode the meaning. The 𝑛2 part of the output makes sense. We can deduce this to
that: (𝑛 + 1)2 − 𝑛2 = 2𝑛 + 1. That means the differences of consecutive square numbers is equal
to 2𝑛 + 1, which is not a random expression. It turns out to be the formula for the odd numbers.
Recall the property that the square numbers can be expressed by the sum of odd numbers. That
means the differences of consecutive squares is equal to a odd number. Hence:

(𝑛 + 1)2 = 𝑛2 + 2𝑛 + 1

We have proved the property!

 Tetrahedron Numbers

What do you think we will get if we sum up the triangular numbers? Of, course we will get a
sequence of numbers, but what geometric shape do you think we will get? Remember you are
adding up triangles of an increasing area…

If you guessed that you would get tetrahedrons you are correct! You will get tetrahedrons of
different volumes, with the volume of the tetrahedron being directly proportional to the value of n.
The tetrahedron numbers are: 1, 4, 10, 20, 35 …
𝑛(𝑛+1)
Here the recursive formula becomes 𝑡𝑛 = 𝑡𝑛−1 + 2
. Since you are basically adding up the
𝑛(𝑛+1)(𝑛+2)
triangular numbers here. The closed formula for this sequence is: 6
, I have derived this
formula used polynomial fitting (see end of the document for the derivation). An interesting
property of these numbers is that if you sum up any two consecutive tetrahedron numbers you get
the sum of consecutive powers. Look at the table below for further detail.

Sum of consecutive numbers Interpretation of the sum

1+4=5 5 = 12 + 22
4 + 10 = 14 14 = 12 + 22 + 32
10 + 20 = 30 30 = 12 + 22 + 32 + 42
20 + 35 = 55 55 = 12 + 22 + 32 + 42 + 52
Using this property we can find a formula for the sum of consecutive squares – a formula that is
useful and important especially in calculus. We can do the same thing we did with the triangular
numbers – we add one to n in the formula and add that to the main formula:
(𝑛 − 1)(𝑛 + 1 − 1)(𝑛 + 2 − 1) (𝑛 − 1)(𝑛)(𝑛 + 1)
=
6 6
𝑛(𝑛 + 1) ∗ (𝑛 + 2) (𝑛 − 1) ∗ (𝑛)(𝑛 + 1)
+
6 6
We can use the distributive property here: 𝑎 ∗ 𝑏 + 𝑎 ∗ 𝑐 = 𝑎(𝑏 + 𝑐) here 𝑎 = 𝑛(𝑛 + 1) and 𝑏 =
(𝑛 − 1) and 𝑐 = (𝑛 + 2). So,

𝑛(𝑛 + 1) ∗ (𝑛 + 2) + (𝑛 − 1) ∗ (𝑛)(𝑛 + 1) = 𝑛(𝑛 + 1)((𝑛 − 1)(𝑛 + 2)) = 𝑛(𝑛 + 1)(2𝑛 + 1)

So, the closed formula for the sum of consecutive squares is:
𝒏(𝒏 + 𝟏)(𝟐𝒏 + 𝟏)
𝟏𝟐 + 𝟐𝟐 + 𝟑𝟐 + 𝟒𝟐 + ⋯ + 𝒏𝟐 =
𝟔
If you see this pattern will continue forever! The partial sum sequence of the tetrahedron numbers
will be: 1, 5, 15, 35, 70 … If you sum up any two consecutive numbers in this sequence you will get
the sum of the sum of the perfect squares!
 Connection to higher dimensional shapes

• Triangular numbers of form triangles when dots are arranged (numbers of dots are given the
by the formula of triangular numbers) Remember this in two dimensions.
• Tetrahedron numbers form tetrahedrons when spheres are arranged (number of spheres
are given by the formula). Remember this is in 3 dimensions.
• The pentalope numbers form pentachoron which is a 4D version of a tetrahedron. number
of spheres are given by the formula). Remember this is in 4 dimensions.