WEBVTT

00:00:05.050 --> 00:00:09.190
In this video I'm going to
explain what is meant by a

00:00:09.190 --> 00:00:12.640
matrix and introduce the
notation that we use when we're

00:00:12.640 --> 00:00:16.780
working with matrices. So let's
start by looking at what we mean

00:00:16.780 --> 00:00:20.575
by a matrix and matrix is a
rectangular pattern of numbers.

00:00:20.575 --> 00:00:21.955
Let's have an example.

00:00:23.440 --> 00:00:25.946
I'm writing down a
pattern of numbers.

00:00:28.790 --> 00:00:33.564
4 -- 113 and 9 and you see they
form a rectangular pattern and

00:00:33.564 --> 00:00:37.656
when we write down a matrix, we
usually enclose the numbers with

00:00:37.656 --> 00:00:41.748
some round brackets like that.
So that's our first example of a

00:00:41.748 --> 00:00:45.499
matrix. Let's have a look at
another example which has a

00:00:45.499 --> 00:00:48.568
different size. So suppose we
have the numbers 12.

00:00:50.080 --> 00:00:54.103
304 and again, this is a
rectangular pattern of

00:00:54.103 --> 00:00:58.573
numbers. I'll put them in
round brackets like that, and

00:00:58.573 --> 00:01:03.043
that's another example of a
matrix. Let's have some more.

00:01:05.560 --> 00:01:07.250
71

00:01:08.560 --> 00:01:12.208
minus three to four and four.

00:01:15.230 --> 00:01:16.590
And a final example.

00:01:18.140 --> 00:01:20.750
A half 00.

00:01:21.840 --> 00:01:28.548
03000 and let's
say nought .7.

00:01:31.480 --> 00:01:35.216
So here we have 4 examples
of matrices.

00:01:37.380 --> 00:01:41.412
They all have different sizes,
so let's look a little bit more

00:01:41.412 --> 00:01:45.780
about how we refer to the size
of a matrix. This first matrix

00:01:45.780 --> 00:01:50.484
here has got two rows and two
columns and we describe it as a

00:01:50.484 --> 00:01:55.188
two by two matrix. We write it
as two by two like that. That's

00:01:55.188 --> 00:01:59.556
just the size. When we write
down the size of a matrix, we

00:01:59.556 --> 00:02:03.252
always give the number of rows
first and the columns second.

00:02:03.252 --> 00:02:05.940
So this has two rows and two
columns.

00:02:07.550 --> 00:02:08.938
What about this matrix?

00:02:09.700 --> 00:02:11.210
This is got one row.

00:02:12.220 --> 00:02:14.959
And 1234 columns.

00:02:16.850 --> 00:02:20.819
So this is a 1 by 4 matrix, one

00:02:20.819 --> 00:02:23.188
row. And four columns.

00:02:25.860 --> 00:02:29.208
This matrix has got 123 rows.

00:02:31.610 --> 00:02:32.909
And two columns.

00:02:35.040 --> 00:02:38.461
So it's a three by two
matrix and the final example

00:02:38.461 --> 00:02:39.705
has got three rows.

00:02:42.610 --> 00:02:46.482
Three columns, so this is a
three by three matrix, so

00:02:46.482 --> 00:02:50.354
remember that we always give
the number of rows first and

00:02:50.354 --> 00:02:51.058
column 2nd.

00:02:52.480 --> 00:02:56.159
The other bit of notation that
we'll need is that we often use

00:02:56.159 --> 00:02:59.555
a capital letter to denote a
matrix, so we might call this

00:02:59.555 --> 00:03:00.687
first matrix here A.

00:03:02.100 --> 00:03:04.188
We might call the second 1B.

00:03:05.380 --> 00:03:06.508
This one C.

00:03:08.020 --> 00:03:09.139
And this 1D.

00:03:10.870 --> 00:03:14.830
So there we are four examples
of matrices, all of different

00:03:14.830 --> 00:03:19.510
sizes and we now know how to
describe a matrix in terms of

00:03:19.510 --> 00:03:23.470
the number of rows and the
number of columns that it

00:03:23.470 --> 00:03:26.710
has. Now that we've seen for
examples of specific

00:03:26.710 --> 00:03:30.670
matrices, let's look at how
we can write down a general

00:03:30.670 --> 00:03:34.270
matrix. Let's suppose this
matrix has the symbol A and

00:03:34.270 --> 00:03:38.950
let's suppose it's got M rows
and N columns, so it's an M

00:03:38.950 --> 00:03:40.030
by N matrix.

00:03:42.820 --> 00:03:47.000
The number that's in the first
row, first column of Matrix

00:03:47.000 --> 00:03:51.560
Capital A will write using a
little A and some subscripts 11

00:03:51.560 --> 00:03:56.880
where the first number refers to
the row label and the 2nd to the

00:03:56.880 --> 00:04:00.300
column label. So this is first
row first column.

00:04:01.730 --> 00:04:06.242
The second number will be a 12,
which corresponds to the first

00:04:06.242 --> 00:04:11.506
row, second column, and so on.
The next one will be a 1 three.

00:04:12.500 --> 00:04:18.467
And so on. Now in this matrix,
because it's an M by N matrix,

00:04:18.467 --> 00:04:24.025
it's got N columns, so the last
number in this first row will be

00:04:24.025 --> 00:04:26.804
a 1 N corresponding to 1st row.

00:04:27.480 --> 00:04:28.510
And column.

00:04:30.990 --> 00:04:35.644
What about the number in here?
Well, it's going to be in the

00:04:35.644 --> 00:04:40.656
2nd row first column, so we'll
call it a 2 one. That's the 2nd

00:04:40.656 --> 00:04:41.730
row, first column.

00:04:42.510 --> 00:04:45.894
The one here will be second
row, second column.

00:04:47.180 --> 00:04:51.674
2nd row, third column, and so on
until we get to the last number

00:04:51.674 --> 00:04:57.225
in this row. Which will be a 2 N
which corresponds to 2nd row,

00:04:57.225 --> 00:05:01.840
NTH column and so on. We can
build up the matrix like this.

00:05:01.840 --> 00:05:06.810
We put all these numbers in as
we want to be 'cause this matrix

00:05:06.810 --> 00:05:08.230
has got M rose.

00:05:08.820 --> 00:05:15.268
The last row here will have a
number AM one corresponding to M

00:05:15.268 --> 00:05:16.756
throw first column.

00:05:18.230 --> 00:05:23.158
NTH Row, second column and so on
all the way along until the last

00:05:23.158 --> 00:05:27.734
number here, which is in the M
throw and the NTH column. So

00:05:27.734 --> 00:05:32.662
we'll call that a MN and that's
the format in which we can write

00:05:32.662 --> 00:05:34.070
down a general matrix.

00:05:35.080 --> 00:05:37.180
Each of these numbers in the

00:05:37.180 --> 00:05:43.300
matrix. We call an element of
the matrix, so A1 one is the

00:05:43.300 --> 00:05:48.030
element in the first row, first
column. In general, the element

00:05:48.030 --> 00:05:54.050
AIJ will be the number that's in
the I throw and the J column.

00:05:55.250 --> 00:05:58.352
Now some of the matrices that
will come across occur so

00:05:58.352 --> 00:06:00.890
frequently or have special
properties that we give them

00:06:00.890 --> 00:06:03.710
special names. Let's have a look
at some of those.

00:06:05.390 --> 00:06:08.978
Let's go back and look again
at the Matrix A. We saw a

00:06:08.978 --> 00:06:09.806
few minutes ago.

00:06:15.320 --> 00:06:17.600
This matrix has got two rows.

00:06:18.290 --> 00:06:20.621
And two columns. So it's a two

00:06:20.621 --> 00:06:25.387
by two matrix. And a matrix
that's got the same number of

00:06:25.387 --> 00:06:29.143
rows and columns like this one,
has we call for obvious reasons

00:06:29.143 --> 00:06:33.306
a square matrix? So this is
the first example of a

00:06:33.306 --> 00:06:33.960
square matrix.

00:06:41.640 --> 00:06:43.890
Now we've already seen another
square matrix because the

00:06:43.890 --> 00:06:46.890
matrix D that we saw a few
minutes ago, which was this

00:06:46.890 --> 00:06:47.140
one.

00:06:56.240 --> 00:06:59.410
He's also a square matrix. This
one's got three rows.

00:07:00.040 --> 00:07:03.670
And three columns. It's a
three by three matrix, and

00:07:03.670 --> 00:07:07.300
because it's got the same
number of rows and columns,

00:07:07.300 --> 00:07:09.115
that's also a square matrix.

00:07:10.960 --> 00:07:14.040
Another term I'd like to
introduce is what's called a

00:07:14.040 --> 00:07:14.656
diagonal matrix.

00:07:21.730 --> 00:07:26.994
If we look again at the matrix
D, we'll see that it has some

00:07:26.994 --> 00:07:30.378
rather special property. This
diagonal, which runs from the

00:07:30.378 --> 00:07:34.514
top left to the bottom right, is
called the leading diagonal.

00:07:41.970 --> 00:07:45.457
And if you look carefully,
you'll see that all the elements

00:07:45.457 --> 00:07:47.359
that are not on the leading

00:07:47.359 --> 00:07:52.832
diagonal are zeros. 0000000 A
matrix for which all the

00:07:52.832 --> 00:07:57.926
elements off the leading
diagonal are zero, is called

00:07:57.926 --> 00:07:59.624
a diagonal matrix.

00:08:01.300 --> 00:08:04.486
There's another special sort of
diagonal matrix I'll introduce

00:08:04.486 --> 00:08:06.610
now. Let's call this one, I.

00:08:10.410 --> 00:08:15.857
Suppose this is a two by two
matrix with ones on the leading

00:08:15.857 --> 00:08:17.533
diagonal and zeros everywhere

00:08:17.533 --> 00:08:20.345
else. So this is a square

00:08:20.345 --> 00:08:24.478
matrix. It's diagonal because
everything off the leading

00:08:24.478 --> 00:08:25.780
diagonal is 0.

00:08:26.400 --> 00:08:29.832
And it's rather special because
on the leading diagonal, all the

00:08:29.832 --> 00:08:34.556
elements are one. Now a matrix
which has this property is

00:08:34.556 --> 00:08:36.024
called an identity matrix.

00:08:37.580 --> 00:08:39.380
Or a unit matrix?

00:08:48.610 --> 00:08:52.040
And when we're working with
matrices, it's usual to reserve

00:08:52.040 --> 00:08:56.156
the letter I for an identity
matrix. Now suppose we have a

00:08:56.156 --> 00:08:57.528
bigger identity matrix. Here's

00:08:57.528 --> 00:09:02.086
another one. Suppose we have a
three by three identity matrix.

00:09:05.080 --> 00:09:07.036
And again, notice that
it's square.

00:09:08.460 --> 00:09:12.276
It's diagonal and there are ones
only on the leading diagonal, so

00:09:12.276 --> 00:09:16.410
this is also an identity matrix.
But because this is a three by

00:09:16.410 --> 00:09:21.180
three and this ones are two by
two and we might not want to mix

00:09:21.180 --> 00:09:25.950
them up and might call this one
I2, because this is a two by two

00:09:25.950 --> 00:09:30.402
matrix and I might call this one
I3. But in both cases these are

00:09:30.402 --> 00:09:33.582
identity matrices and we'll see
that identity matrices have a

00:09:33.582 --> 00:09:37.080
very important role to play when
we look at matrix multiplication

00:09:37.080 --> 00:09:38.352
in a forthcoming video.