WEBVTT

00:00:06.860 --> 00:00:09.560
PROFESSOR: Hi, welcome
back to recitation.

00:00:09.560 --> 00:00:13.490
In the lecture, we've learned
very important concepts--

00:00:13.490 --> 00:00:16.620
linear space and
linear subspace.

00:00:16.620 --> 00:00:20.200
Well, as you can imagine, if
we call something a space,

00:00:20.200 --> 00:00:24.670
we're putting a lot of things,
a lot of objects into one set.

00:00:24.670 --> 00:00:28.400
But for linear space, we
want to put them in according

00:00:28.400 --> 00:00:30.680
to a particular manner.

00:00:30.680 --> 00:00:33.890
So can you recall what are
the conditions for a set

00:00:33.890 --> 00:00:36.280
to be a linear space?

00:00:36.280 --> 00:00:40.490
You take any two elements from
that set, take the sum of them.

00:00:40.490 --> 00:00:43.400
You want the sum to be
still in the same set.

00:00:43.400 --> 00:00:45.370
That's the first condition.

00:00:45.370 --> 00:00:50.010
Second, you take any multiple
of any element from that set,

00:00:50.010 --> 00:00:53.190
the result will
still be in that set.

00:00:53.190 --> 00:00:56.140
That's the second condition.

00:00:56.140 --> 00:00:58.970
And if, within a
linear space, you

00:00:58.970 --> 00:01:03.510
can find a subset which also
satisfies the two conditions,

00:01:03.510 --> 00:01:06.560
that will give you a subspace.

00:01:06.560 --> 00:01:09.620
Today, we're going to
look at this example

00:01:09.620 --> 00:01:13.880
to review these two important
properties of linear space

00:01:13.880 --> 00:01:16.870
and subspace.

00:01:16.870 --> 00:01:20.980
I have two vectors, x_1 and x_2.

00:01:20.980 --> 00:01:23.480
Both of them are vectors in R^3.

00:01:23.480 --> 00:01:26.910
So, as you can see,
I've drawn them here.

00:01:26.910 --> 00:01:30.750
This is x_1, and this is x_2.

00:01:30.750 --> 00:01:36.740
So first, we want to find the
subspace generated by x_1.

00:01:36.740 --> 00:01:38.820
I call it V_1.

00:01:38.820 --> 00:01:43.430
Let me say a word about
this "generated by."

00:01:43.430 --> 00:01:48.190
So what do I mean by a
subspace generated by x_1?

00:01:48.190 --> 00:01:53.200
I'm looking for the smallest
subspace that contains x_1, as

00:01:53.200 --> 00:01:55.470
small as you can get.

00:01:55.470 --> 00:02:00.690
Similarly, I want to find out
the subspace generated by x_2,

00:02:00.690 --> 00:02:03.520
call it V_2.

00:02:03.520 --> 00:02:07.830
Then we want to say something
about the intersection of V_1

00:02:07.830 --> 00:02:09.240
and V_2.

00:02:09.240 --> 00:02:11.080
That's the first question.

00:02:11.080 --> 00:02:16.570
And the second question, we
want to put vector x_1 and x_2

00:02:16.570 --> 00:02:21.040
together and look at the
subspace generated by x_1

00:02:21.040 --> 00:02:22.890
and x_2 at the same time.

00:02:22.890 --> 00:02:25.950
So I call it V_3.

00:02:25.950 --> 00:02:28.400
And a good question
to be asked here

00:02:28.400 --> 00:02:34.820
is: what is the relation
of v3 to V_1 union V_2?

00:02:34.820 --> 00:02:37.830
Do you think they're equal?

00:02:37.830 --> 00:02:40.320
Within the second
question, I would also

00:02:40.320 --> 00:02:45.850
like you to find a subspace,
call it S, of V_3 such

00:02:45.850 --> 00:02:51.730
that neither x_1
nor x_2 is in s.

00:02:51.730 --> 00:02:54.330
And the last
question is I'd like

00:02:54.330 --> 00:02:58.030
you to say something about
the intersection of V_3

00:02:58.030 --> 00:03:00.380
with the xy-plane.

00:03:00.380 --> 00:03:04.850
So notice that, of course,
xy-plane is also a subspace

00:03:04.850 --> 00:03:06.170
of R^3.

00:03:06.170 --> 00:03:11.740
So again, I'm looking at the
intersection of two subspaces.

00:03:11.740 --> 00:03:14.170
All right, why don't
you hit the pause now,

00:03:14.170 --> 00:03:17.860
and try to solve these
three problems on your own.

00:03:17.860 --> 00:03:23.040
And I'd like you to identify
your answers in this picture

00:03:23.040 --> 00:03:24.980
whenever you can.

00:03:24.980 --> 00:03:27.740
I'll come back later and
continue working with you.

00:03:36.620 --> 00:03:39.280
OK, how did your drawing go?

00:03:39.280 --> 00:03:41.350
Let's look at this together.

00:03:41.350 --> 00:03:45.660
First, we want to find
subspace generated by x_1.

00:03:45.660 --> 00:03:47.520
So here is x_1.

00:03:47.520 --> 00:03:50.100
Let's keep in mind
the two conditions

00:03:50.100 --> 00:03:52.710
that a subspace has to satisfy.

00:03:52.710 --> 00:03:57.080
Well, if I want to
obtain the subspace,

00:03:57.080 --> 00:04:01.210
at least I have to be able
to take any multiple of x_1,

00:04:01.210 --> 00:04:01.880
right?

00:04:01.880 --> 00:04:07.320
So that means at least I have to
include the straight line that

00:04:07.320 --> 00:04:09.000
contains x_1.

00:04:09.000 --> 00:04:13.340
So I'm going to try to draw
the straight line here.

00:04:13.340 --> 00:04:17.430
So I'm just simply going
to extend this x1 along

00:04:17.430 --> 00:04:19.384
to both directions.

00:04:19.384 --> 00:04:32.520
So x_1, I will try very
hard to make it straight.

00:04:32.520 --> 00:04:35.400
But to be honest, it's
really hard for me

00:04:35.400 --> 00:04:39.020
to draw straight
lines on the board.

00:04:39.020 --> 00:04:40.315
Is that straight?

00:04:40.315 --> 00:04:41.990
Seems fine.

00:04:41.990 --> 00:04:46.690
All right, so this
entire line contains x_1.

00:04:46.690 --> 00:04:50.982
So at least this line
has to be in V_1.

00:04:50.982 --> 00:04:55.000
Is there anything
else beyond this line?

00:04:55.000 --> 00:04:57.920
Now let's turn to
the second condition.

00:04:57.920 --> 00:04:59.990
The second condition
says that we

00:04:59.990 --> 00:05:04.110
have to be able to take any
two elements, take the sum,

00:05:04.110 --> 00:05:06.500
and the sum will
remain in that set.

00:05:06.500 --> 00:05:08.440
Does it work for this line?

00:05:08.440 --> 00:05:12.030
You take any two
vectors on this line,

00:05:12.030 --> 00:05:14.940
or you can say any two
points on this line,

00:05:14.940 --> 00:05:17.140
and then you take
to sum of them.

00:05:17.140 --> 00:05:20.450
Of course, you still get
something on this line.

00:05:20.450 --> 00:05:23.340
You won't be able
to escape from it.

00:05:23.340 --> 00:05:26.910
Which means this line
is a perfect set that

00:05:26.910 --> 00:05:29.220
satisfies the two conditions.

00:05:29.220 --> 00:05:33.880
So this line simply
gives me V_1.

00:05:33.880 --> 00:05:37.010
That is the smallest
subspace that contains x_1.

00:05:37.010 --> 00:05:42.580
So in other words, the subspace
generated by x_1 is V_1.

00:05:42.580 --> 00:05:45.560
So similarly, let's look at x_2.

00:05:45.560 --> 00:05:48.700
What is the subspace
generated by x_2?

00:05:48.700 --> 00:05:51.030
Again, you get the
entire straight line

00:05:51.030 --> 00:05:53.050
that contains x_2.

00:05:53.050 --> 00:05:57.445
So I'm going to extend
x_2 in both directions.

00:06:08.530 --> 00:06:09.490
I hope it's straight.

00:06:12.480 --> 00:06:14.650
Not too bad.

00:06:14.650 --> 00:06:16.810
That will give me V_2.

00:06:16.810 --> 00:06:19.840
All right, V_1, V_2.

00:06:19.840 --> 00:06:23.890
Both of them are
subspaces of R^3.

00:06:23.890 --> 00:06:27.630
Now let's look at the
intersection of V_1 and V_2.

00:06:27.630 --> 00:06:31.280
So we know that both of
them are straight lines.

00:06:31.280 --> 00:06:33.690
And clearly they're
not parallel,

00:06:33.690 --> 00:06:36.770
because x_1 and x_2
are not parallel.

00:06:36.770 --> 00:06:40.650
So what is the intersection
of V_1 and V_2?

00:06:40.650 --> 00:06:47.520
The intersection of V_1
and V_2 is the only point

00:06:47.520 --> 00:06:49.110
at which they cross.

00:06:49.110 --> 00:06:51.490
And where's that point?

00:06:51.490 --> 00:06:53.280
It's here, right at the origin.

00:06:53.280 --> 00:06:56.420
Because both of them
pass the origin.

00:06:56.420 --> 00:06:59.690
I'm going to use O to denote.

00:06:59.690 --> 00:07:03.360
That's the intersection
of V_1 and V_2.

00:07:03.360 --> 00:07:07.120
This is a set with
only one element.

00:07:07.120 --> 00:07:10.250
What can you say about this set?

00:07:10.250 --> 00:07:15.330
I claim this is also
a subspace of R^3.

00:07:15.330 --> 00:07:20.210
By saying space, we usually
mean a lot of objects together.

00:07:20.210 --> 00:07:22.110
But look at this set.

00:07:22.110 --> 00:07:24.580
This set fits perfectly
into the conditions

00:07:24.580 --> 00:07:27.040
of being a linear space.

00:07:27.040 --> 00:07:33.260
You take any multiple of 0,
again you get 0-- 0 plus 0,

00:07:33.260 --> 00:07:34.470
you get 0.

00:07:34.470 --> 00:07:38.840
So that's a perfectly
fine subspace of R^3.

00:07:38.840 --> 00:07:41.340
All right, so what
we have got here is:

00:07:41.340 --> 00:07:44.940
I take the intersection
of V_1 and V_2,

00:07:44.940 --> 00:07:48.510
and the result, again,
becomes a subspace,

00:07:48.510 --> 00:07:51.230
which only contains the origin.

00:07:51.230 --> 00:07:53.640
That completes the
first question.

00:07:53.640 --> 00:07:56.210
Let's look at the second one.

00:07:56.210 --> 00:08:00.980
In the second question, I want
to put x_1 and x_2 together,

00:08:00.980 --> 00:08:05.860
and look at the subspace
generated by x_1 and x_2.

00:08:05.860 --> 00:08:08.790
And I would also like
you to say something

00:08:08.790 --> 00:08:13.140
about the relation
between V_1 union V_2

00:08:13.140 --> 00:08:17.210
to the subspace
generated by x_1, x_2.

00:08:17.210 --> 00:08:20.980
Let's try to answer the
second question first.

00:08:20.980 --> 00:08:26.100
Is there a chance that
V_1 union V_2 equal to v3?

00:08:26.100 --> 00:08:29.220
OK, so what is V_1 union V_2?

00:08:29.220 --> 00:08:31.700
That's clearly just
two lines, right?

00:08:31.700 --> 00:08:35.360
This line union this line.

00:08:35.360 --> 00:08:39.710
Is there a chance that this
union will be a subspace?

00:08:39.710 --> 00:08:41.830
Let's check the two conditions.

00:08:41.830 --> 00:08:47.260
First, you take any multiple
of the elements in this union.

00:08:47.260 --> 00:08:52.210
It's either on this
line or on this line.

00:08:52.210 --> 00:08:56.080
Seems that the multiple is still
going to stay inside the union.

00:08:56.080 --> 00:08:58.596
So the first condition
is actually satisfied.

00:08:58.596 --> 00:09:00.700
What about the second one?

00:09:00.700 --> 00:09:02.760
The second one says
that I have to be

00:09:02.760 --> 00:09:07.280
able to take any sum, the
sum of any two elements,

00:09:07.280 --> 00:09:09.240
from these two lines.

00:09:09.240 --> 00:09:12.610
Let's just try a simple
sum, x_1 plus x_2.

00:09:12.610 --> 00:09:14.210
So what is x_1 plus x_2?

00:09:18.570 --> 00:09:20.710
You just sum up each coordinate.

00:09:20.710 --> 00:09:26.370
That will give you [2, 5, 3].

00:09:29.500 --> 00:09:32.880
In this picture-- can I
draw it in this picture?

00:09:32.880 --> 00:09:41.045
It's going to be somewhere here.

00:09:50.280 --> 00:09:54.824
That's x_1 plus x_3.

00:09:54.824 --> 00:09:57.560
Did you notice something?

00:09:57.560 --> 00:10:02.530
You clearly have got
out of this union.

00:10:02.530 --> 00:10:11.230
So this sum is not
inside V_1 union V_2.

00:10:11.230 --> 00:10:15.730
Which means V_1 union
V-2 is not a subspace.

00:10:15.730 --> 00:10:20.830
Then it's impossible that
this union will equal to V_3.

00:10:20.830 --> 00:10:24.640
So the answer to
the second question

00:10:24.640 --> 00:10:29.020
is no, V_3 is not
equal to V_1 union V_2.

00:10:29.020 --> 00:10:31.540
Now let's identify V_3.

00:10:31.540 --> 00:10:35.340
Well, as you can see, since we
have seen from this argument

00:10:35.340 --> 00:10:38.650
that we have to
be able to include

00:10:38.650 --> 00:10:40.490
this diagonal vector here.

00:10:40.490 --> 00:10:44.170
But in fact, as you can see
since we can take any elements

00:10:44.170 --> 00:10:46.400
from these two
lines, we're actually

00:10:46.400 --> 00:10:53.680
including every vector on the
plane spanned by V_1 and V_2.

00:10:53.680 --> 00:11:01.460
So in other words, I'm actually
looking at this huge plane

00:11:01.460 --> 00:11:06.690
that is spanned by V_1 and V_2.

00:11:06.690 --> 00:11:08.520
That will give me V_3.

00:11:12.600 --> 00:11:14.300
So that's reasonable.

00:11:14.300 --> 00:11:18.380
I'm looking at the subspace
generated by two lines in R^3.

00:11:18.380 --> 00:11:23.780
And that two line will be
able to span a plane in R^3.

00:11:23.780 --> 00:11:27.780
Now for the last
part of question two,

00:11:27.780 --> 00:11:30.710
I want to find a
subspace S of V_3--

00:11:30.710 --> 00:11:33.210
so I want to find a
subspace of this plane--

00:11:33.210 --> 00:11:38.460
such that x_1 is not S,
x_2 is not in S either.

00:11:38.460 --> 00:11:43.530
So I want to stay away from
this line and this line.

00:11:43.530 --> 00:11:47.880
Can you find such
a subspace of V_3?

00:11:47.880 --> 00:11:50.470
It's right here.

00:11:50.470 --> 00:11:55.280
Because if you look at this
vector, x_1 plus x_2-- sorry,

00:11:55.280 --> 00:11:56.300
that should be x_2.

00:12:00.240 --> 00:12:02.960
If you look at this
vector, and if you

00:12:02.960 --> 00:12:06.520
look at the subspace
generated by this vector,

00:12:06.520 --> 00:12:08.990
again you know it's
going to be a line.

00:12:08.990 --> 00:12:10.390
And this line is right here.

00:12:18.790 --> 00:12:22.840
This line forms a
perfect subspace of V_3.

00:12:22.840 --> 00:12:27.960
But neither x_1 nor x_2
is inside this subspace.

00:12:27.960 --> 00:12:33.640
So let's just make it S. Of
course, the choice is not

00:12:33.640 --> 00:12:34.480
unique.

00:12:34.480 --> 00:12:40.280
You can take twice x_1 plus
x_2 or x_1 plus twice x_2.

00:12:40.280 --> 00:12:44.350
OK, we have completed
the second problem.

00:12:44.350 --> 00:12:49.380
The last problems ask us to
find the intersection of V_3

00:12:49.380 --> 00:12:51.430
with xy-plane.

00:12:51.430 --> 00:12:52.930
Well, just think about that.

00:12:52.930 --> 00:12:56.350
We've identified
V_3 as this plane

00:12:56.350 --> 00:12:59.500
spanned by line V_1 and V_2.

00:12:59.500 --> 00:13:02.150
And xy is also a plane.

00:13:02.150 --> 00:13:05.290
So we're talking about the
intersection of two planes

00:13:05.290 --> 00:13:06.830
in R^3.

00:13:06.830 --> 00:13:08.660
What would that be?

00:13:08.660 --> 00:13:11.590
So you have two
planes intersect.

00:13:11.590 --> 00:13:15.120
The intersection will be a
straight line again, right?

00:13:15.120 --> 00:13:18.080
So let's locate
that straight line.

00:13:18.080 --> 00:13:23.050
We want to find something that
is inside V_3 and xy-plane

00:13:23.050 --> 00:13:23.850
at the same time.

00:13:26.480 --> 00:13:30.380
What can you say about
the points in xy-plane?

00:13:30.380 --> 00:13:33.880
The z-coordinate has to be 0.

00:13:33.880 --> 00:13:38.720
And at the same time, we know
that at least x_1 and x_2

00:13:38.720 --> 00:13:40.290
are in V_3.

00:13:40.290 --> 00:13:43.130
So did you notice that?

00:13:43.130 --> 00:13:47.760
x_2 is a vector
that lies in V_3.

00:13:47.760 --> 00:13:52.830
But at the same time, the
z-coordinate of x_2 is 0.

00:13:52.830 --> 00:13:55.160
That is what we're looking for.

00:13:55.160 --> 00:14:09.550
So the intersection
of V_3 with xy-plane

00:14:09.550 --> 00:14:15.090
will simply be the
line that contains x_2.

00:14:15.090 --> 00:14:20.820
And here we've
identified that as V_2.

00:14:20.820 --> 00:14:22.340
That's it.

00:14:22.340 --> 00:14:24.610
This is a subspace of R^3.

00:14:24.610 --> 00:14:26.580
And this is the subspace of R^3.

00:14:26.580 --> 00:14:32.740
The intersection is
again a subspace of R^3.

00:14:32.740 --> 00:14:36.590
I hope you've learned a
way to somehow visualize

00:14:36.590 --> 00:14:41.380
this linear space and subspace
through this exercise.

00:14:41.380 --> 00:14:44.407
Thank you for watching,
and I'll see you next time.