Skip to main content

Vector Summation with Gauss's Formula

·379 words·2 mins
Author
Adrian

Vector Summation Trick (Inspired by Gauss)
#

In this article we’re going to be dealing with a little vector sum.

We’re given:

\[ \sum_{i=1}^{100} \begin{bmatrix} i - 1 \\ i \\ i + 1 \end{bmatrix} \]

And if we don’t recognize this sum immediately, it’s usually a good idea to try a few terms and look for a pattern.


Trying a Few Terms
#

When \( i = 1 \):

\[ \begin{bmatrix} 0 \\ 1 \\ 2 \end{bmatrix} \]

When \( i = 2 \):

\[ \begin{bmatrix} 1 \\ 2 \\ 3 \end{bmatrix} \]

When \( i = 3 \):

\[ \begin{bmatrix} 2 \\ 3 \\ 4 \end{bmatrix} \]

… and so on, until \( i = 100 \):

\[ \begin{bmatrix} 99 \\ 100 \\ 101 \end{bmatrix} \]

So, what are we really summing here?


Adding Vectors
#

When you add column vectors, you add component-wise. So this becomes:

[ \sum_{i=1}^{100} \begin{bmatrix} i - 1 \ i \ i + 1 \end{bmatrix}
#

\begin{bmatrix} \sum_{i=1}^{100} (i - 1) \ \sum_{i=1}^{100} i \ \sum_{i=1}^{100} (i + 1) \end{bmatrix} ]

Now we can break each of these down.


Middle Term: Gauss Trick
#

Everyone knows how to add:

\[ 1 + 2 + 3 + \dots + 100 \]

And for that, there’s a trick supposedly discovered by Gauss in elementary school.

Pair numbers from the ends:

  • \( 1 + 100 = 101 \)
  • \( 2 + 99 = 101 \)
  • \( 3 + 98 = 101 \)

There are 50 such pairs:

\[ \sum_{i=1}^{100} i = 50 \times 101 = 5050 \]

So the middle element in our vector sum is:

\[ \sum_{i=1}^{100} i = 5050 \]

Top Term: Each \( i - 1 \)
#

This is just shifting each term down by 1, so it’s:

\[ \sum_{i=1}^{100} (i - 1) = \sum_{i=1}^{100} i - 100 = 5050 - 100 = 4950 \]

Bottom Term: Each \( i + 1 \)
#

Each term is bumped up by 1:

\[ \sum_{i=1}^{100} (i + 1) = \sum_{i=1}^{100} i + 100 = 5050 + 100 = 5150 \]

Final Answer
#

So, the entire sum simplifies to:

\[ \begin{bmatrix} 4950 \\ 5050 \\ 5150 \end{bmatrix} \]

Credits
#

All explanations and problems written by Adrian Hernandez Vega, unless otherwise noted.

Prefer video? Watch the full explanation here: