WEBVTT

00:00:00.526 --> 00:00:01.573
- [Voiceover] Hello everyone.

00:00:01.573 --> 00:00:03.683
So what I'd like to do here is to describe

00:00:03.683 --> 00:00:06.220
how we think about
three-dimensional graphs.

00:00:06.220 --> 00:00:08.858
Three-dimensional graphs
are a way that we represent

00:00:08.858 --> 00:00:10.255
certain kind of multi-variable function

00:00:10.255 --> 00:00:12.156
that kind of has two inputs,

00:00:12.156 --> 00:00:14.539
or rather a two-dimensional input,

00:00:14.539 --> 00:00:17.003
and then one-dimensional
of output of some kind.

00:00:17.003 --> 00:00:19.302
So the one that I have pictured here

00:00:19.302 --> 00:00:23.733
is f of (x, y) equals x
squared plus y squared.

00:00:23.733 --> 00:00:26.618
And before talking
exactly about this graph,

00:00:26.618 --> 00:00:28.463
I think it would be helpful, by analogy,

00:00:28.463 --> 00:00:30.380
we take a look at the
two-dimensional graphs and

00:00:30.380 --> 00:00:32.580
kinda remind ourselves how those work,

00:00:32.580 --> 00:00:36.562
what it is that we do,
because, it's pretty much

00:00:36.562 --> 00:00:38.876
the same thing in three-dimensions,

00:00:38.876 --> 00:00:41.148
but it takes a little bit
more of visualization.

00:00:41.148 --> 00:00:43.365
So the two-dimensional graphs,

00:00:43.365 --> 00:00:45.559
they have some kind of function, you know,

00:00:45.559 --> 00:00:49.726
let's see you have f of
x is equal to x squared,

00:00:51.086 --> 00:00:54.000
and anytime you visualizing
a function, you trying to

00:00:54.000 --> 00:00:55.554
understand the relationship between

00:00:55.554 --> 00:00:57.154
the inputs and the outputs.

00:00:57.154 --> 00:00:59.032
And here those are both just numbers,

00:00:59.032 --> 00:01:00.873
so you know you input a number like two,

00:01:00.873 --> 00:01:03.646
and it's gonna output four,

00:01:03.646 --> 00:01:07.351
you know you input negative
one it's gonna output one.

00:01:07.351 --> 00:01:09.638
And you're trying to
understand all the possible

00:01:09.638 --> 00:01:11.506
input-output pairs.

00:01:11.506 --> 00:01:13.089
And the fact that we can do this,

00:01:13.089 --> 00:01:17.571
that we can get a pretty
good intuitive feel for

00:01:17.571 --> 00:01:21.376
every possible input-output
pair is pretty incredible,

00:01:21.376 --> 00:01:24.476
the way we go about this
with graphs is you think

00:01:24.476 --> 00:01:26.492
we just plotting these
actual pairs, right?

00:01:26.492 --> 00:01:30.189
So you're gonna plot the
point, let's say we are gonna

00:01:30.189 --> 00:01:34.306
plot the point (2,4), so we
may kind of mark our graph,

00:01:34.306 --> 00:01:36.973
two here, one, two, three, four,

00:01:39.221 --> 00:01:43.261
so you wanna mark somewhere here (2,4),

00:01:43.261 --> 00:01:45.862
and that represents an input-output pair.

00:01:45.862 --> 00:01:48.300
And if you do that with,
you know, negative one, one,

00:01:48.300 --> 00:01:50.383
you go negative one, one.

00:01:51.586 --> 00:01:55.302
And when you do this for every
possible input-output pair,

00:01:55.302 --> 00:01:59.159
what you end up getting, I
might not draw this super well,

00:01:59.159 --> 00:02:01.491
is some kind of smooth curve.

00:02:01.491 --> 00:02:04.232
The implication for doing this
is that we typically think of

00:02:04.232 --> 00:02:07.406
what is on the x-axis as
being where the inputs live,

00:02:07.406 --> 00:02:09.813
you know, this would be, we
think of as the input one,

00:02:09.813 --> 00:02:13.018
and this is the input two, and so on,

00:02:13.018 --> 00:02:17.185
and then you think of the
output as being the height

00:02:18.828 --> 00:02:21.554
of the graph above each point.

00:02:21.554 --> 00:02:23.272
But this is kind of a
consequence of the fact

00:02:23.272 --> 00:02:26.078
where we just listing
all of the pairs here.

00:02:26.078 --> 00:02:29.149
Now if we go to the world
of multi-variable functions,

00:02:29.149 --> 00:02:31.905
you know, not gonna show
the graph right now,

00:02:31.905 --> 00:02:34.122
let's just think we've got
three-dimensional space

00:02:34.122 --> 00:02:37.106
at out disposal to do with what we will.

00:02:37.106 --> 00:02:39.010
We still want to understand
the relationship between

00:02:39.010 --> 00:02:42.653
inputs and outputs of
this guy, but this case,

00:02:42.653 --> 00:02:46.820
inputs are something that we
think of as pair of points,

00:02:47.914 --> 00:02:51.391
we might have a pair of points like (1,2),

00:02:51.391 --> 00:02:53.654
and the output there is gonna be

00:02:53.654 --> 00:02:57.821
one squared plus two squared,
and that equals is five.

00:03:02.573 --> 00:03:05.328
So how do we visualize that?

00:03:05.328 --> 00:03:08.053
Well if we wanna pair these
things together, the natural way

00:03:08.053 --> 00:03:10.813
to do that is to think of
a triplet of some kind.

00:03:10.813 --> 00:03:14.980
So in this case, you wanna
plug the triplet (1, 2, 5),

00:03:17.268 --> 00:03:19.932
and to do that in three-dimensions,

00:03:19.932 --> 00:03:23.157
we'll take a look over
here, we think of going one

00:03:23.157 --> 00:03:25.938
in the x direction, this
axis here is the x-axis,

00:03:25.938 --> 00:03:28.276
so we want to move distance one there,

00:03:28.276 --> 00:03:31.700
and we want to go two in the y direction,

00:03:31.700 --> 00:03:34.610
so we kinda think of
going distance two there,

00:03:34.610 --> 00:03:37.442
and then five up, and then

00:03:37.442 --> 00:03:40.120
that's gonna give us some
kind of point, right?

00:03:40.120 --> 00:03:41.489
So we think this point in space

00:03:41.489 --> 00:03:44.043
and that's a given input-output pair.

00:03:44.043 --> 00:03:45.807
But we could do this for a lot, right,

00:03:45.807 --> 00:03:48.571
a couple different
points that you might get

00:03:48.571 --> 00:03:51.183
if you start plotting
various different ones,

00:03:51.183 --> 00:03:54.014
look something like this,
and of course there is

00:03:54.014 --> 00:03:56.244
infinitely many that you can
do and it'll take forever

00:03:56.244 --> 00:03:58.926
if you try to just draw each
one in three-dimensions,

00:03:58.926 --> 00:04:02.042
but what's really nice here
is that you know get rid of

00:04:02.042 --> 00:04:04.567
those lines, if you imagine doing this

00:04:04.567 --> 00:04:07.289
for all of the infinite
many pairs of inputs

00:04:07.289 --> 00:04:11.892
that you could possibly have,
you end up drawing a surface.

00:04:11.892 --> 00:04:14.952
So in this case the surface
kind of looks like a

00:04:14.952 --> 00:04:17.215
three-dimensional parabola,
that's no coincidence,

00:04:17.215 --> 00:04:19.038
it has to do with the
fact that we are using

00:04:19.038 --> 00:04:21.331
x squared and y squared here.

00:04:21.331 --> 00:04:25.498
And now the inputs like (1,
2), we think of as being

00:04:27.177 --> 00:04:29.301
on the xy-plane, right?

00:04:29.301 --> 00:04:31.658
So you think of the inputs living here,

00:04:31.658 --> 00:04:33.818
and then what corresponds
to the output is that

00:04:33.818 --> 00:04:37.463
height of a giving point
above the graph, right?

00:04:37.463 --> 00:04:39.564
So it's very similar to
two-dimensions, you think,

00:04:39.564 --> 00:04:42.084
you know, we think of the
inputs as being on one axis,

00:04:42.084 --> 00:04:44.417
and the height gives the output there.

00:04:44.417 --> 00:04:45.936
So just to give an example of

00:04:45.936 --> 00:04:49.283
what the consequence of this
is, I want you to think about

00:04:49.283 --> 00:04:52.522
what might happen if we change
our multi-variable function

00:04:52.522 --> 00:04:56.690
a little bit, and we multiply
everything by half, right?

00:04:56.690 --> 00:05:00.857
So I'll draw in red here, let's
see that we have a function,

00:05:02.178 --> 00:05:06.452
but I'm gonna change it so
that it outputs one half

00:05:06.452 --> 00:05:09.227
of x squared plus y squared.

00:05:09.227 --> 00:05:13.394
What's gonna be the shape of
the graph for that function?

00:05:14.286 --> 00:05:15.919
And what it means is the
height of every point

00:05:15.919 --> 00:05:19.645
above this xy-plane is gonna
have to get cut in half.

00:05:19.645 --> 00:05:21.392
So it's actually just the modification

00:05:21.392 --> 00:05:23.420
of what we already have,
but everything kind of

00:05:23.420 --> 00:05:27.181
sloops on down to be
about half of what it was.

00:05:27.181 --> 00:05:29.386
So in this case instead
of that height being five,

00:05:29.386 --> 00:05:31.531
it'll be two-point-five.

00:05:31.531 --> 00:05:33.442
You could imagine, let's
say we did this, you know,

00:05:33.442 --> 00:05:35.823
is even more extreme,
instead of saying one-half,

00:05:35.823 --> 00:05:38.725
you cut it down by like one-twelfth,

00:05:38.725 --> 00:05:42.558
maybe I'll use the same
color, by one-twelfth,

00:05:43.440 --> 00:05:45.017
that would mean that everything, you know,

00:05:45.017 --> 00:05:49.259
sloops very flat, very flat
and close to the xy-plane.

00:05:49.259 --> 00:05:51.443
So the graph being very
close to xy-plane like this

00:05:51.443 --> 00:05:54.909
corresponds to very small outputs.

00:05:54.909 --> 00:05:57.287
And one thing that I'd like
to caution you against,

00:05:57.287 --> 00:05:59.563
it's very tempting to try to think of

00:05:59.563 --> 00:06:01.520
every multi-variable function as a graph,

00:06:01.520 --> 00:06:03.687
cause we are so used to
graphs in two-dimensions

00:06:03.687 --> 00:06:06.726
and we are so used to
trying to find analogies

00:06:06.726 --> 00:06:09.781
between two-dimensions and
three-dimensions directly,

00:06:09.781 --> 00:06:12.734
but the only reason that
this works is because

00:06:12.734 --> 00:06:15.376
if you take the number of
dimension in the input,

00:06:15.376 --> 00:06:17.515
two-dimensions, and then
the number of dimensions

00:06:17.515 --> 00:06:19.926
in the output, one-dimension,
it was reasonable

00:06:19.926 --> 00:06:23.281
to fit all of that into
three, which we could do.

00:06:23.281 --> 00:06:25.104
But imagine if you have
a multi-variable function

00:06:25.104 --> 00:06:27.044
with, you know, a three-dimensional input,

00:06:27.044 --> 00:06:29.156
and a two-dimensional
output, that would require

00:06:29.156 --> 00:06:31.420
a five-dimensional graph,
but we are not very good

00:06:31.420 --> 00:06:33.899
at visualizing things like that.

00:06:33.899 --> 00:06:35.586
So there's lots of other
methods, and I think

00:06:35.586 --> 00:06:37.883
it's very important to
kinda of open you mind

00:06:37.883 --> 00:06:39.752
to what those might be.

00:06:39.752 --> 00:06:42.940
In particular, another one
that I'm gonna go through soon,

00:06:42.940 --> 00:06:44.613
lets us think about 3-D graphs but kind of

00:06:44.613 --> 00:06:46.709
in a two-dimensional setting,
and we are just gonna

00:06:46.709 --> 00:06:50.093
look at the input space,
that's called a contour map.

00:06:50.093 --> 00:06:52.194
Couple of other ones,
like parametric functions,

00:06:52.194 --> 00:06:54.216
you just look in the output space;

00:06:54.216 --> 00:06:55.616
things like vector space,

00:06:55.616 --> 00:06:59.210
you kind of look at the input
space but get all the outputs.

00:06:59.210 --> 00:07:00.770
There's lots of different
ways, I'll go over those

00:07:00.770 --> 00:07:03.301
in the next few videos.

00:07:03.301 --> 00:07:05.409
And that's three-dimensional graphs.