1
00:00:00 --> 00:00:06

2
00:00:04 --> 00:00:10
The topic for today is how to
change variables.

3
00:00:08 --> 00:00:14
So, we're talking about
substitutions and differential

4
00:00:13 --> 00:00:19
equations, or changing
variables.

5
00:00:16 --> 00:00:22
That might seem like a sort of
fussy thing to talk about in the

6
00:00:22 --> 00:00:28
third or fourth lecture,
but the reason is that so far,

7
00:00:27 --> 00:00:33
you know how to solve two kinds
of differential equations,

8
00:00:32 --> 00:00:38
two kinds of first-order
differential equations,

9
00:00:36 --> 00:00:42
one where you can separate
variables, and the linear

10
00:00:41 --> 00:00:47
equation that we talked about
last time.

11
00:00:47 --> 00:00:53
Now, the sad fact is that in
some sense, those are the only

12
00:00:51 --> 00:00:57
two general methods there are,
that those are the only two

13
00:00:56 --> 00:01:02
kinds of equations that can
always be solved.

14
00:01:00 --> 00:01:06
Well, what about all the
others?

15
00:01:02 --> 00:01:08
The answer is that to a great
extent, all the other equations

16
00:01:06 --> 00:01:12
that can be solved,
the solution can be done by

17
00:01:09 --> 00:01:15
changing the variables in the
equation to reduce it to one of

18
00:01:14 --> 00:01:20
the cases that we can already
do.

19
00:01:16 --> 00:01:22
Now, I'm going to give you two
examples of that,

20
00:01:19 --> 00:01:25
two significant examples of
that today.

21
00:01:22 --> 00:01:28
But, ultimately,
as you will see,

22
00:01:24 --> 00:01:30
the way the equations are
solved is by changing them into

23
00:01:28 --> 00:01:34
a linear equation,
or an equation where the

24
00:01:31 --> 00:01:37
variables are separable.
However, that's for a few

25
00:01:36 --> 00:01:42
minutes.
The first change of variables

26
00:01:39 --> 00:01:45
that I want to talk about is an
almost trivial one.

27
00:01:43 --> 00:01:49
But it's the most common kind
there is, and you've already had

28
00:01:47 --> 00:01:53
it in physics class.
But I think it's so important

29
00:01:51 --> 00:01:57
in the science and engineering
subjects that it's a good idea,

30
00:01:55 --> 00:02:01
even in 18.03,
to call attention to it

31
00:01:58 --> 00:02:04
explicitly.
So, in that sense,

32
00:02:01 --> 00:02:07
the most common change of
variables is the one simple one

33
00:02:06 --> 00:02:12
called scaling.
So, again, the kind of equation

34
00:02:11 --> 00:02:17
I'm talking about is a general
first-order equation.

35
00:02:15 --> 00:02:21
And, scaling simply means to
change the coordinates,

36
00:02:20 --> 00:02:26
in effect, or axes,
to change the coordinates on

37
00:02:24 --> 00:02:30
the axes to scale the axes,
to either stretch them or

38
00:02:29 --> 00:02:35
contract them.
So, what does the change of

39
00:02:34 --> 00:02:40
variable actually look like?
Well, it means you introduce

40
00:02:39 --> 00:02:45
new variables,
where x1 is equal to x times

41
00:02:42 --> 00:02:48
something or times a constant.
I'll write it as divided by a

42
00:02:48 --> 00:02:54
constant, since that tends to be
a little bit more the way people

43
00:02:53 --> 00:02:59
think of it.
And y, the same.

44
00:02:56 --> 00:03:02
So, the new variable y1 is
related to the old one by an

45
00:03:00 --> 00:03:06
equation of that form.
So, a, b are constants.

46
00:03:06 --> 00:03:12
So, those are the equations.
Why does one do this?

47
00:03:12 --> 00:03:18
Well, for a lot of reasons.
But, maybe we can list them.

48
00:03:19 --> 00:03:25
You, for example,
could be changing units.

49
00:03:24 --> 00:03:30
That's a common reason in
physics.

50
00:03:28 --> 00:03:34
Changing the units that he
used, you would have to make a

51
00:03:35 --> 00:03:41
change of coordinates of this
form.

52
00:03:41 --> 00:03:47
Perhaps the even more important
reason is to,

53
00:03:45 --> 00:03:51
sometimes it's used to make the
variables dimensionless.

54
00:03:50 --> 00:03:56
In other words,
so that the variables become

55
00:03:55 --> 00:04:01
pure numbers,
with no units attached to them.

56
00:03:59 --> 00:04:05
Since you are well aware of the
tortures involved in dealing

57
00:04:05 --> 00:04:11
with units in physics,
the point of making variables,

58
00:04:10 --> 00:04:16
I'm sorry, dimensionless,
I don't have to sell that.

59
00:04:17 --> 00:04:23
Dimensionless,
i.e.

60
00:04:18 --> 00:04:24
no units, without any units
attached.

61
00:04:21 --> 00:04:27
It just represents the number
three, not three seconds,

62
00:04:26 --> 00:04:32
or three grahams,
or anything like that.

63
00:04:30 --> 00:04:36
And, the third reason is to
reduce or simplify the

64
00:04:34 --> 00:04:40
constants: reduce the number or
simplify the constants in the

65
00:04:40 --> 00:04:46
equation.
Reduce their number is self

66
00:04:44 --> 00:04:50
explanatory.
Simplify means make them less,

67
00:04:48 --> 00:04:54
either dimensionless also,
or if you can't do that,

68
00:04:53 --> 00:04:59
at least less dependent upon
the critical units than the old

69
00:04:59 --> 00:05:05
ones were.
Let me give you a very simple

70
00:05:04 --> 00:05:10
example which will illustrate
most of these things.

71
00:05:08 --> 00:05:14
It's the equation.
It's a version of the cooling

72
00:05:12 --> 00:05:18
law, which applies at very high
temperatures,

73
00:05:16 --> 00:05:22
and it runs.
So, it's like Newton's cooling

74
00:05:20 --> 00:05:26
laws, except it's the internal
and external temperatures vary,

75
00:05:25 --> 00:05:31
what's important is not the
first power as in Newton's Law,

76
00:05:30 --> 00:05:36
but the fourth power.
So, it's a constant.

77
00:05:34 --> 00:05:40
And, the difference is,
now, it's the external

78
00:05:38 --> 00:05:44
temperature, which,
just so there won't be so many

79
00:05:42 --> 00:05:48
capital T's in the equation,
I'm going to call M,

80
00:05:46 --> 00:05:52
to the forth power minus T to
the forth power.

81
00:05:50 --> 00:05:56
So, T is the internal

82
00:05:53 --> 00:05:59
temperature, the thing we are
interested in.

83
00:05:57 --> 00:06:03
And, M is the external
constant, which I'll assume,

84
00:06:01 --> 00:06:07
now, is a constant external
temperature.

85
00:06:06 --> 00:06:12
So, this is valid if big
temperature differences,

86
00:06:10 --> 00:06:16
Newton's Law,
breaks down and one needs a

87
00:06:14 --> 00:06:20
different one.
Now, you are free to solve that

88
00:06:19 --> 00:06:25
equation just as it stands,
if you can.

89
00:06:23 --> 00:06:29
There are difficulties
connected with it because you're

90
00:06:28 --> 00:06:34
dealing with the fourth powers,
of course.

91
00:06:34 --> 00:06:40
But, before you do that,
one should scale.

92
00:06:37 --> 00:06:43
How should I scale?
Well, I'm going to scale by

93
00:06:42 --> 00:06:48
relating T to M.
So, that is very likely to use

94
00:06:46 --> 00:06:52
is T1 equals T divided by M.

95
00:06:50 --> 00:06:56
This is now dimensionless
because M, of course,

96
00:06:55 --> 00:07:01
has the units of temperature,
degrees Celsius,

97
00:06:59 --> 00:07:05
degrees absolute,
whatever it is,

98
00:07:02 --> 00:07:08
as does T.
And therefore,

99
00:07:06 --> 00:07:12
by taking the ratio of the two,
there are no units attached to

100
00:07:11 --> 00:07:17
it.
So, this is dimensionless.

101
00:07:14 --> 00:07:20
Now, how actually do I change
the variable in the equation?

102
00:07:19 --> 00:07:25
Well, watch this.
It's an utterly trivial idea,

103
00:07:23 --> 00:07:29
and utterly important.
Don't slog around doing it this

104
00:07:27 --> 00:07:33
way, trying to stuff it in,
and divide first.

105
00:07:31 --> 00:07:37
Instead, do the inverse.
In other words,

106
00:07:36 --> 00:07:42
write it instead as T equals
MT1, the reason being that it's

107
00:07:41 --> 00:07:47
T that's facing you in that
equation, and therefore T you

108
00:07:46 --> 00:07:52
want to substitute for.
So, let's do it.

109
00:07:49 --> 00:07:55
The new equation will be what?
Well, dT-- Since this is a

110
00:07:54 --> 00:08:00
constant, the left-hand side
becomes dT1 / dt times M equals

111
00:07:59 --> 00:08:05
k times M to the forth minus M
to the forth T1 to the forth,

112
00:08:04 --> 00:08:10
so I'm going to factor

113
00:08:09 --> 00:08:15
out that M to the forth,
and make it one minus T1 to the

114
00:08:14 --> 00:08:20
forth, okay?

115
00:08:18 --> 00:08:24
Now, I could divide through by
M and get rid of one of those,

116
00:08:23 --> 00:08:29
and so, the new equation,
now, is dT1 / dt,

117
00:08:27 --> 00:08:33
d time, is equal to-- Now,
I have k M cubed out

118
00:08:32 --> 00:08:38
front here.
I'm going to just give that a

119
00:08:36 --> 00:08:42
new name, k1.
Essentially,

120
00:08:39 --> 00:08:45
it's the same equation.
It's no harder to solve and no

121
00:08:44 --> 00:08:50
easier to solve than the
original one.

122
00:08:47 --> 00:08:53
But it's been simplified.
For one, I think it looks

123
00:08:52 --> 00:08:58
better.
So, to compare the two,

124
00:08:55 --> 00:09:01
I'll put this one up in green,
and this one in green,

125
00:09:00 --> 00:09:06
too, just to convince you it's
the same, but indicate that it's

126
00:09:06 --> 00:09:12
the same equation.
Notice, so, T1 has been

127
00:09:11 --> 00:09:17
rendered, is now dimensionless.
So, I don't have to even ask

128
00:09:16 --> 00:09:22
when I solve this equation,
oh, please tell me what the

129
00:09:20 --> 00:09:26
units of temperature are.
How you are measuring

130
00:09:24 --> 00:09:30
temperature makes no difference
to this equation.

131
00:09:27 --> 00:09:33
k1 still has units.
What units does it have?

132
00:09:32 --> 00:09:38
It's been simplified because it
now has the units of,

133
00:09:37 --> 00:09:43
since this is dimensionless and
this is dimensionless,

134
00:09:42 --> 00:09:48
it has the units of inverse
time.

135
00:09:45 --> 00:09:51
So, k1, whereas it had units
involving both degrees and

136
00:09:51 --> 00:09:57
seconds before,
now it has inverse time as its

137
00:09:55 --> 00:10:01
units.
And, moreover,

138
00:09:57 --> 00:10:03
there's one less constant.
So, one less constant in the

139
00:10:02 --> 00:10:08
equation.
It just looks better.

140
00:10:07 --> 00:10:13
This business,
I think you know that k1,

141
00:10:11 --> 00:10:17
the process of forming k1 out
of k M cubed is

142
00:10:18 --> 00:10:24
called lumping constants.
I think they use standard

143
00:10:23 --> 00:10:29
terminology in physics and
engineering courses.

144
00:10:30 --> 00:10:36
Try to get all the constants
together like this.

145
00:10:33 --> 00:10:39
And then you lump them there.
They are lumped for you,

146
00:10:36 --> 00:10:42
and then you just give the lump
a new name.

147
00:10:39 --> 00:10:45
So, that's an example of
scaling.

148
00:10:41 --> 00:10:47
Watch out for when you can use
that.

149
00:10:43 --> 00:10:49
For example,
it would have probably been a

150
00:10:46 --> 00:10:52
good thing to use in the first
problem set when you were

151
00:10:50 --> 00:10:56
handling this problem of drug
elimination and hormone

152
00:10:53 --> 00:10:59
elimination production inside of
the thing.

153
00:10:56 --> 00:11:02
You could lump constants,
and as was done to some extent

154
00:11:00 --> 00:11:06
on the solutions to get a neater
looking answer,

155
00:11:03 --> 00:11:09
one without so many constants
in it.

156
00:11:07 --> 00:11:13
Okay, let's now go to serious
stuff, where we are actually

157
00:11:11 --> 00:11:17
going to make changes of
variables which we hope will

158
00:11:16 --> 00:11:22
render unsolvable equations
suddenly solvable.

159
00:11:20 --> 00:11:26
Now, I'm going to do that by
making substitutions.

160
00:11:24 --> 00:11:30
But, it's, I think,
quite important to watch up

161
00:11:28 --> 00:11:34
there are two kinds of
substitutions.

162
00:11:31 --> 00:11:37
There are direct substitutions.
That's where you introduce a

163
00:11:37 --> 00:11:43
new variable.
I don't know how to write this

164
00:11:41 --> 00:11:47
on the board.
I'll just write it

165
00:11:43 --> 00:11:49
schematically.
So, it's one which says that

166
00:11:47 --> 00:11:53
the new variable is equal to
some combination of the old

167
00:11:52 --> 00:11:58
variables.
The other kind of substitution

168
00:11:56 --> 00:12:02
is inverse.
It's just the reverse.

169
00:12:00 --> 00:12:06
Here, you say that the old
variables are some combination

170
00:12:04 --> 00:12:10
of the new.
Now, often you'll have to stick

171
00:12:07 --> 00:12:13
in a few old variables,
too.

172
00:12:09 --> 00:12:15
But the basic,
it's what appears on the

173
00:12:12 --> 00:12:18
left-hand side.
Is it a new variable that

174
00:12:15 --> 00:12:21
appears on the left-hand side by
itself, or is it the old

175
00:12:19 --> 00:12:25
variable that appears on the
left-hand side?

176
00:12:23 --> 00:12:29
Now, right here,
we have an example.

177
00:12:25 --> 00:12:31
If I did it as a direct
substitution,

178
00:12:28 --> 00:12:34
I would have written T1 equals
T over M.

179
00:12:34 --> 00:12:40
That's the way I define the new
variable, which of course you

180
00:12:38 --> 00:12:44
have to do if you're introducing
it.

181
00:12:41 --> 00:12:47
But when I actually did the
substitution,

182
00:12:44 --> 00:12:50
I did the inverse substitution.
Namely, I used T equals T1,

183
00:12:48 --> 00:12:54
M times T1. And,

184
00:12:50 --> 00:12:56
the reason for doing that was
because it was the capital T's

185
00:12:55 --> 00:13:01
that faced me in the equation
and I had to have something to

186
00:12:59 --> 00:13:05
replace them with.
Now, you see this already in

187
00:13:03 --> 00:13:09
calculus, this distinction.
But that might have been a year

188
00:13:07 --> 00:13:13
and a half ago.
Just let me remind you,

189
00:13:10 --> 00:13:16
typically in calculus,
for example,

190
00:13:12 --> 00:13:18
when you want to do this kind
of integral, let's say,

191
00:13:15 --> 00:13:21
x times the square root of one
minus x squared dx,

192
00:13:19 --> 00:13:25
the substitution you'd use for

193
00:13:23 --> 00:13:29
that is u equals one minus x
squared,

194
00:13:26 --> 00:13:32
right?
And then, you calculate,

195
00:13:28 --> 00:13:34
and then you would observe that
this, the x dx,

196
00:13:31 --> 00:13:37
more or less makes up du,
apart from a constant factor,

197
00:13:35 --> 00:13:41
there.
So, this would be an example of

198
00:13:39 --> 00:13:45
direct substitution.
You put it in and convert the

199
00:13:42 --> 00:13:48
integral into an integral of u.
What would be an example of

200
00:13:46 --> 00:13:52
inverse substitution?
Well, if I take away the x and

201
00:13:50 --> 00:13:56
ask you, instead,
to do this integral,

202
00:13:52 --> 00:13:58
then you know that the right
thing to do is not to start with

203
00:13:56 --> 00:14:02
u, but to start with the x and
write x equals sine or cosine u.

204
00:14:02 --> 00:14:08
So, this is a direct
substitution in that integral,

205
00:14:05 --> 00:14:11
but this integral calls for an
inverse substitution in order to

206
00:14:09 --> 00:14:15
be able to do it.
And notice, they would look

207
00:14:12 --> 00:14:18
practically the same.
But, of course,

208
00:14:15 --> 00:14:21
as you know from your
experience, they are not.

209
00:14:18 --> 00:14:24
They're very different.
Okay, so I'm going to watch for

210
00:14:21 --> 00:14:27
that distinction as I do these
examples.

211
00:14:24 --> 00:14:30
The first one I want to do is
an example as a direct

212
00:14:28 --> 00:14:34
substitution.

213
00:14:30 --> 00:14:36

214

215

216
00:14:47 --> 00:14:53
So, it applies to the equation
of the form y prime equals,

217
00:14:52 --> 00:14:58
there are two kinds of terms on
the right-hand side.

218
00:14:56 --> 00:15:02
Let's use p of x,
p of x times y plus q of x

219
00:15:00 --> 00:15:06
times any power 
whatsoever of y.

220
00:15:05 --> 00:15:11
Well, notice,
for example,

221
00:15:07 --> 00:15:13
if n were zero,
what kind of equation would

222
00:15:11 --> 00:15:17
this be?
y to the n would be

223
00:15:14 --> 00:15:20
one, and this would be a linear
equation, which you know how to

224
00:15:20 --> 00:15:26
solve.
So, n equals zero we already

225
00:15:23 --> 00:15:29
know how to do.
So, let's assume that n is not

226
00:15:27 --> 00:15:33
zero, so that we're in new
territory.

227
00:15:33 --> 00:15:39
Well, if n were equal to one,
you could separate variables.

228
00:15:38 --> 00:15:44
So, that too is not exciting.
But, nonetheless,

229
00:15:42 --> 00:15:48
it will be included in what I'm
going to say now.

230
00:15:46 --> 00:15:52
If n is two or three,
or n could be one half.

231
00:15:50 --> 00:15:56
So anything:
even zero is all right.

232
00:15:53 --> 00:15:59
It's just silly.
Any number: it could be

233
00:15:57 --> 00:16:03
negative.
n equals minus five.

234
00:15:59 --> 00:16:05
That would be fine also.
This kind of equation,

235
00:16:04 --> 00:16:10
to give it its name,
is called the Bernoulli

236
00:16:08 --> 00:16:14
equation, named after which
Bernoulli, I haven't the

237
00:16:11 --> 00:16:17
faintest idea.
There were, I think,

238
00:16:14 --> 00:16:20
three or four of them.
And, they fought with each

239
00:16:18 --> 00:16:24
other.
But, they were all smart.

240
00:16:20 --> 00:16:26
Now, the key trick,
if you like,

241
00:16:23 --> 00:16:29
method, to solving any
Bernoulli equation,

242
00:16:26 --> 00:16:32
let me call another thing.
Most important is what's

243
00:16:30 --> 00:16:36
missing.
It must not have a pure x term

244
00:16:34 --> 00:16:40
in it.
And that goes for a constant

245
00:16:37 --> 00:16:43
term.
In other words,

246
00:16:38 --> 00:16:44
it must look exactly like this.
Everything multiplied by y,

247
00:16:43 --> 00:16:49
or a power of y,
two terms.

248
00:16:45 --> 00:16:51
So, for example,
if I add one to this,

249
00:16:48 --> 00:16:54
the equation becomes
non-doable.

250
00:16:51 --> 00:16:57
Right, it's very easy to
contaminate it into an equation

251
00:16:55 --> 00:17:01
that's unsolvable.
It's got to look just like

252
00:16:59 --> 00:17:05
that.
Now, you've got one on your

253
00:17:03 --> 00:17:09
homework.
You've got several.

254
00:17:05 --> 00:17:11
Both part one and part two have
Bernoulli equations on them.

255
00:17:10 --> 00:17:16
So, this is practical,
in some sense.

256
00:17:13 --> 00:17:19
What do we got?
The idea is to divide by y to

257
00:17:17 --> 00:17:23
the n.
Ignore all formulas that you're

258
00:17:20 --> 00:17:26
given.
Just remember that when you see

259
00:17:23 --> 00:17:29
something that looks like this,
or something that you can turn

260
00:17:28 --> 00:17:34
into something that looks like
this, divide through by y to the

261
00:17:34 --> 00:17:40
nth power, no matter what n is.
All right, so y prime over y to

262
00:17:40 --> 00:17:46
the n is equal to p of x times
one over y to the n minus one,

263
00:17:44 --> 00:17:50
right, plus q of x.

264
00:17:49 --> 00:17:55
Well, that certainly doesn't
look any better than what I

265
00:17:53 --> 00:17:59
started with.
And, in your terms,

266
00:17:55 --> 00:18:01
it probably looks somewhat
worse because it's got all those

267
00:17:59 --> 00:18:05
Y's at the denominator,
and who wants to see them

268
00:18:03 --> 00:18:09
there?
But, look at it.

269
00:18:06 --> 00:18:12
In this very slightly
transformed Bernoulli equation

270
00:18:10 --> 00:18:16
is a linear equation struggling
to be free.

271
00:18:14 --> 00:18:20
Where is it?
Why is it trying to be a linear

272
00:18:17 --> 00:18:23
equation?
Make a new variable,

273
00:18:20 --> 00:18:26
call this hunk of it in new
variable.

274
00:18:23 --> 00:18:29
Let's call it V.
So, V is equal to one over y to

275
00:18:27 --> 00:18:33
the n minus one.

276
00:18:30 --> 00:18:36
Or, if you like,
you can think of that as y to

277
00:18:34 --> 00:18:40
the one minus n.
What's V prime?

278
00:18:39 --> 00:18:45
So, this is the direct
substitution I am going to use,

279
00:18:44 --> 00:18:50
but of course,
the problem is,

280
00:18:46 --> 00:18:52
what am I going to use on this?
Well, the little miracle

281
00:18:51 --> 00:18:57
happens.
What's the derivative of this?

282
00:18:54 --> 00:19:00
It is one minus n times y to
the negative n times y prime

283
00:18:59 --> 00:19:05
In other words,

284
00:19:04 --> 00:19:10
up to a constant,
this constant factor,

285
00:19:07 --> 00:19:13
one minus n,
it's exactly the left-hand side

286
00:19:11 --> 00:19:17
of the equation.
Well, let's make N not equal

287
00:19:15 --> 00:19:21
one either.
As I said, you could separate

288
00:19:18 --> 00:19:24
variables if n equals one.
What's the equation,

289
00:19:22 --> 00:19:28
then, turned into?
A Bernoulli equation,

290
00:19:28 --> 00:19:34
divided through in this way,
is then turned into the

291
00:19:36 --> 00:19:42
equation one minus n,
sorry, V prime divided by one

292
00:19:44 --> 00:19:50
minus n is equal to p of x times
V plus q of x.

293
00:19:55 --> 00:20:01
It's linear.

294
00:20:01 --> 00:20:07
And now, solve it as a linear
equation.

295
00:20:03 --> 00:20:09
Solve it as a linear equation.
You notice, it's not in

296
00:20:06 --> 00:20:12
standard form,
not in standard linear form.

297
00:20:09 --> 00:20:15
To do that, you're going to
have to put the p on the other

298
00:20:13 --> 00:20:19
side.
That's okay,

299
00:20:14 --> 00:20:20
that term, on the other side,
solve it, and at the end,

300
00:20:17 --> 00:20:23
don't forget that you put in
the V.

301
00:20:19 --> 00:20:25
It wasn't in the original
problem.

302
00:20:22 --> 00:20:28
So, you have to convert the
problem, the answer,

303
00:20:25 --> 00:20:31
back in terms of y.
It'll come out in terms of V,

304
00:20:28 --> 00:20:34
but you must put it back in
terms of y.

305
00:20:32 --> 00:20:38
Let's do a really simple
example just to illustrate the

306
00:20:38 --> 00:20:44
method, and to illustrate the
fact that I don't want you to

307
00:20:45 --> 00:20:51
memorize formulas.
Learn methods,

308
00:20:49 --> 00:20:55
not final formulas.
So, suppose the equation is,

309
00:20:54 --> 00:21:00
let's say, y prime equals y
over x minus y squared.

310
00:21:01 --> 00:21:07
That's a Bernoulli equation.

311
00:21:06 --> 00:21:12
I could, of course,
have concealed it by writing xy

312
00:21:09 --> 00:21:15
prime plus xy prime minus xy
equals negative y squared.

313
00:21:13 --> 00:21:19
Then, it wouldn't look
instantly like a Bernoulli

314
00:21:16 --> 00:21:22
equation.
You would have to stare at it a

315
00:21:19 --> 00:21:25
while and say,
hey, that's a Bernoulli

316
00:21:22 --> 00:21:28
equation.
Okay, but so I'm handing it to

317
00:21:25 --> 00:21:31
you a silver platter,
as it were.

318
00:21:27 --> 00:21:33
So, what do we do?
Divide through by y squared.

319
00:21:32 --> 00:21:38
So, it's y prime over y squared
equals one over x times one over

320
00:21:38 --> 00:21:44
y minus one.

321
00:21:41 --> 00:21:47
And now, the substitution,
then, I'm going to make,

322
00:21:46 --> 00:21:52
is for this thing.
V equals one over y.

323
00:21:51 --> 00:21:57
It's a direct substitution.

324
00:21:53 --> 00:21:59
V prime is going to be negative
one over y squared

325
00:21:59 --> 00:22:05
times y prime.
Don't forget to use the chain

326
00:22:05 --> 00:22:11
rule when you differentiate with
respect-- because the

327
00:22:08 --> 00:22:14
differentiation is with respect
to x, of course,

328
00:22:12 --> 00:22:18
not with respect to y.
Okay, so what's this thing?

329
00:22:15 --> 00:22:21
That's the left-hand side.
The only thing is it's got a

330
00:22:19 --> 00:22:25
negative sign.
So, this is minus V prime

331
00:22:22 --> 00:22:28
equals, one over x stays one
over x, one over y.

332
00:22:26 --> 00:22:32
So, it's V over x minus one.

333
00:22:30 --> 00:22:36
So, let's put that in standard
form.

334
00:22:32 --> 00:22:38
In standard form,
it will look like,

335
00:22:35 --> 00:22:41
first imagine multiplying it
through by negative one,

336
00:22:39 --> 00:22:45
and then putting this term on
the other side.

337
00:22:42 --> 00:22:48
And, it will turn into V prime
plus V over X is equal to one.

338
00:22:47 --> 00:22:53
So, that's the linear equation

339
00:22:51 --> 00:22:57
in standard linear form that we
are asked to solve.

340
00:22:54 --> 00:23:00
And, the solution isn't very
hard.

341
00:22:57 --> 00:23:03
The integrating factor is,
well, I integrate one over x.

342
00:23:03 --> 00:23:09
That makes log x.
And, e to the log x,

343
00:23:05 --> 00:23:11
so, it's e to the log x,
which is, of course,

344
00:23:09 --> 00:23:15
just x itself.
So, I should multiply this

345
00:23:12 --> 00:23:18
through by x squared,
be able to integrate it.

346
00:23:15 --> 00:23:21
Now, some of you,
I would hope,

347
00:23:17 --> 00:23:23
just can see that right away,
that if you multiply this

348
00:23:21 --> 00:23:27
through by x,
it's going to look good.

349
00:23:24 --> 00:23:30
So, after we multiply through
by x, which I get?

350
00:23:27 --> 00:23:33
(xV) prime for the-- maybe I
shouldn't skip a step.

351
00:23:33 --> 00:23:39
You are still learning this
stuff, so let's not skip a step.

352
00:23:38 --> 00:23:44
So, it becomes x V prime plus V
equals x,

353
00:23:44 --> 00:23:50
okay?
After I multiplied through by

354
00:23:47 --> 00:23:53
the integrating factor,
this now says this is xV prime,

355
00:23:53 --> 00:23:59
and I quickly check that that,
in fact, is what it's equal to,

356
00:23:59 --> 00:24:05
equals x, and therefore xV is
equal to one half x squared plus

357
00:24:05 --> 00:24:11
a constant. And,

358
00:24:08 --> 00:24:14
therefore, V is equal to one
half x plus C over x.

359
00:24:14 --> 00:24:20
You can leave it at that form,

360
00:24:19 --> 00:24:25
or you can combine terms.
It doesn't matter much.

361
00:24:23 --> 00:24:29
Am I done?
The answer is,

362
00:24:25 --> 00:24:31
no I am not done,
because nobody reading this

363
00:24:29 --> 00:24:35
answer would know what V was.
V wasn't in the original

364
00:24:33 --> 00:24:39
problem.
It was y that was in the

365
00:24:35 --> 00:24:41
original problem.
And therefore,

366
00:24:37 --> 00:24:43
the relation is,
one is the reciprocal of the

367
00:24:40 --> 00:24:46
other.
And therefore,

368
00:24:41 --> 00:24:47
I have to turn this expression
upside down.

369
00:24:44 --> 00:24:50
Well, if you're going to have
to turn it upside down,

370
00:24:47 --> 00:24:53
you probably should make it
look a little better.

371
00:24:50 --> 00:24:56
Let's rewrite it as x squared
plus 2c,

372
00:24:53 --> 00:24:59
combining fractions,
I think they call it in high

373
00:24:56 --> 00:25:02
school or elementary school,
plus 2c.

374
00:25:00 --> 00:25:06
How's that? x squared 
plus 2c divided by 2x.

375
00:25:03 --> 00:25:09
Now, 2c, you don't call it

376
00:25:07 --> 00:25:13
constant 2c because this is just
as arbitrary to call it c1.

377
00:25:12 --> 00:25:18
So, I'll call that,
so, my answer will be y equals

378
00:25:16 --> 00:25:22
2x divided by x squared plus an
arbitrary constant.

379
00:25:20 --> 00:25:26
But, to indicate it's different
from that one,

380
00:25:23 --> 00:25:29
I'll call it C1. C1 is

381
00:25:27 --> 00:25:33
two times the old one,
but that doesn't really matter.

382
00:25:31 --> 00:25:37
So, there's the solution.
It has an arbitrary constant in

383
00:25:37 --> 00:25:43
it, but you note it's not an
additive arbitrary constant.

384
00:25:40 --> 00:25:46
The arbitrary constant is
tucked into the solution.

385
00:25:44 --> 00:25:50
If you had to satisfy an
initial condition,

386
00:25:47 --> 00:25:53
you would take this form,
and starting from this form,

387
00:25:50 --> 00:25:56
figure out what C1 was in order
to satisfy that initial

388
00:25:54 --> 00:26:00
condition.
Thus, Bernoulli equation is

389
00:25:57 --> 00:26:03
solved.
Our first Bernoulli equation:

390
00:25:59 --> 00:26:05
isn't that exciting?
So, here was the equation,

391
00:26:05 --> 00:26:11
and there is its solution.
Now, the one I'm asking you to

392
00:26:11 --> 00:26:17
solve on the problem set in part
two is no harder than this,

393
00:26:18 --> 00:26:24
except I ask you some hard
questions about it,

394
00:26:24 --> 00:26:30
not very hard,
but a little hard about it.

395
00:26:30 --> 00:26:36
I hope you will find them
interesting questions.

396
00:26:33 --> 00:26:39
You already have the
experimental evidence from the

397
00:26:37 --> 00:26:43
first problem set,
and the problem is to explain

398
00:26:40 --> 00:26:46
the experimental evidence by
actually solving the equation in

399
00:26:45 --> 00:26:51
the scene.
I think you'll find it

400
00:26:47 --> 00:26:53
interesting.
But, maybe that's just a pious

401
00:26:51 --> 00:26:57
hope.
Okay, I like,

402
00:26:52 --> 00:26:58
now, to turn to the second
method, where a second class of

403
00:26:56 --> 00:27:02
equations which require inverse
substitution,

404
00:27:00 --> 00:27:06
and those are equations,
which are called homogeneous,

405
00:27:04 --> 00:27:10
a highly overworked word in
differential equations,

406
00:27:08 --> 00:27:14
and in mathematics in general.
But, it's unfortunately just

407
00:27:14 --> 00:27:20
the right word to describe them.
So, these are homogeneous,

408
00:27:19 --> 00:27:25
first-order ODE's.
Now, I already used the word in

409
00:27:23 --> 00:27:29
one context in talking about the
linear equations when zero is

410
00:27:28 --> 00:27:34
the right hand side.
This is different,

411
00:27:32 --> 00:27:38
but nonetheless,
the two uses of the word have

412
00:27:35 --> 00:27:41
the same common source.
The homogeneous differential

413
00:27:39 --> 00:27:45
equation, homogeneous newspeak,
is y prime equals,

414
00:27:43 --> 00:27:49
it's a question of what the
right hand side looks like.

415
00:27:47 --> 00:27:53
And, now, the supposed way to
say it is, you should be able to

416
00:27:52 --> 00:27:58
write the right-hand side as a
function of a combined variable,

417
00:27:57 --> 00:28:03
y divided by x.
In other words,

418
00:28:01 --> 00:28:07
after fooling around with the
right hand side a little bit,

419
00:28:06 --> 00:28:12
you should be able to write it
so that every time a variable

420
00:28:11 --> 00:28:17
appears, it's always in the
combination y over x.

421
00:28:15 --> 00:28:21
Let me give some examples.
For example,

422
00:28:19 --> 00:28:25
suppose y prime were,
let's say, x squared y divided

423
00:28:23 --> 00:28:29
by x cubed plus y cubed.

424
00:28:29 --> 00:28:35
Well, that doesn't look in that
form.

425
00:28:31 --> 00:28:37
Well, yes it is.
Imagine dividing the top and

426
00:28:34 --> 00:28:40
bottom by x cubed.
What would you get?

427
00:28:37 --> 00:28:43
The top would be y over x,
if you divided it by x

428
00:28:40 --> 00:28:46
cubed.
And, if I divide the bottom by

429
00:28:43 --> 00:28:49
x cubed, also,
which, of course,

430
00:28:45 --> 00:28:51
doesn't change the value of the
fraction, as they say in

431
00:28:49 --> 00:28:55
elementary school,
one plus (y over x) cubed.

432
00:28:52 --> 00:28:58
So, this is the way it started

433
00:28:55 --> 00:29:01
out looking, but you just said
ah-ha, that was a homogeneous

434
00:28:59 --> 00:29:05
equation because I could see it
could be written that way.

435
00:29:05 --> 00:29:11
How about another homogeneous
equation?

436
00:29:10 --> 00:29:16
How about x y prime?
Is that a homogeneous equation?

437
00:29:18 --> 00:29:24
Of course it is:
otherwise, why would I be

438
00:29:24 --> 00:29:30
talking about it?
If you divide through by x,

439
00:29:29 --> 00:29:35
you can tuck it inside the
radical, the square root,

440
00:29:33 --> 00:29:39
if you remember to square it
when you do that.

441
00:29:36 --> 00:29:42
And, it becomes the square root
of x squared over x squared,

442
00:29:41 --> 00:29:47
which is one,
plus y squared over x squared.

443
00:29:44 --> 00:29:50
It's homogeneous.

444
00:29:47 --> 00:29:53
Now, you might say,
hey, this looks like you had to

445
00:29:50 --> 00:29:56
be rather clever to figure out
if an equation is homogeneous.

446
00:29:55 --> 00:30:01
Is there some other way?
Yeah, there is another way,

447
00:29:58 --> 00:30:04
and it's the other way which
explains why it's called

448
00:30:02 --> 00:30:08
homogeneous.
You can think of it this way.

449
00:30:07 --> 00:30:13
It's an equation which is,
in modern speak,

450
00:30:12 --> 00:30:18
invariant, invariant under the
operation zoom.

451
00:30:18 --> 00:30:24
What is zoom?
Zoom is, you increase the scale

452
00:30:23 --> 00:30:29
equally on both axes.
So, the zoom operation is the

453
00:30:30 --> 00:30:36
one which sends x into 
a times x,

454
00:30:36 --> 00:30:42
and y into a times y.

455
00:30:42 --> 00:30:48
In other words,
you change the scale on both

456
00:30:46 --> 00:30:52
axes by the same factor,
a.

457
00:30:48 --> 00:30:54
Now, what I say is,
gee, maybe you shouldn't write

458
00:30:53 --> 00:30:59
it like this.
Why don't we say,

459
00:30:56 --> 00:31:02
we introduce,
how about this?

460
00:31:00 --> 00:31:06
So, think of it as a change of
variables.

461
00:31:02 --> 00:31:08
We will write it like that.
So, you can put here an equals

462
00:31:06 --> 00:31:12
sign, if you don't know what
this meaningless arrow means.

463
00:31:10 --> 00:31:16
So, I'm making this change of
variables, and I'm describing it

464
00:31:14 --> 00:31:20
as an inverse substitution.
But of course,

465
00:31:16 --> 00:31:22
it wouldn't make any
difference.

466
00:31:19 --> 00:31:25
It's exactly the same as the
direct substitution I started

467
00:31:22 --> 00:31:28
out with underscaling.
The only difference is,

468
00:31:25 --> 00:31:31
I'm not using different scales
on both axes.

469
00:31:28 --> 00:31:34
I'm expanding them both
equally.

470
00:31:32 --> 00:31:38
That's what I mean by zoom.
Now, what happens to the

471
00:31:36 --> 00:31:42
equation?
Well, what happens to dy over

472
00:31:40 --> 00:31:46
dx?
Well, dx is a dx1.

473
00:31:43 --> 00:31:49
dy is a dy1.

474
00:31:47 --> 00:31:53
And therefore,
the ratio, dy by dx is the same

475
00:31:51 --> 00:31:57
as dy1 over dx1.

476
00:31:54 --> 00:32:00
So, the left-hand side becomes
dy1 over dx1,

477
00:31:58 --> 00:32:04
and the right-hand side becomes
F of, well, y over x is the same

478
00:32:04 --> 00:32:10
as y over, since I've scaled
them equally,

479
00:32:08 --> 00:32:14
this is the same as y1 over x1.

480
00:32:14 --> 00:32:20
So, it's y1 over x1,
and the net effect is I simply,

481
00:32:18 --> 00:32:24
everywhere I have an x,
I change it to x1,

482
00:32:22 --> 00:32:28
and wherever I have a y,
I change it to y1,

483
00:32:25 --> 00:32:31
which, what's in a name?
It's the identical equation.

484
00:32:31 --> 00:32:37
So, I haven't changed the
equation at all via zoom

485
00:32:35 --> 00:32:41
transformation.
And, that's what makes it

486
00:32:38 --> 00:32:44
homogeneous.
That's not an uncommon use of

487
00:32:42 --> 00:32:48
the word homogeneous.
When you say space is

488
00:32:45 --> 00:32:51
homogeneous, every direction,
well, that means,

489
00:32:49 --> 00:32:55
I don't know.
It means, okay,

490
00:32:51 --> 00:32:57
I'm getting into trouble there.
I'll let it go since I can't

491
00:32:56 --> 00:33:02
prepare any better,
I haven't prepared any better

492
00:33:00 --> 00:33:06
explanation, but this is a
pretty good one.

493
00:33:06 --> 00:33:12
Okay, so, suppose we've got a
homogeneous equation.

494
00:33:13 --> 00:33:19
How do we solve it?
So, here's our equation,

495
00:33:20 --> 00:33:26
F of y over x.
Well, what substitution would

496
00:33:29 --> 00:33:35
you like to make?
Obviously, we should make a

497
00:33:34 --> 00:33:40
direct substitution,
z equals y over x.

498
00:33:38 --> 00:33:44
So, why did he say that this
was going to be an example of

499
00:33:42 --> 00:33:48
inverse substitution?
Because I wanted to confuse

500
00:33:45 --> 00:33:51
you.
But look, that's fine.

501
00:33:47 --> 00:33:53
If you write it in that form,
you'll know exactly what to do

502
00:33:51 --> 00:33:57
with the right-hand side.
And, this is why everybody

503
00:33:55 --> 00:34:01
loves to do that.
But for Charlie,

504
00:33:57 --> 00:34:03
you have to substitute into the
left-hand side as well.

505
00:34:03 --> 00:34:09
And, I can testify,
for many years of looking with

506
00:34:06 --> 00:34:12
sinking heart at examination
papers, what happens if you try

507
00:34:10 --> 00:34:16
to make a direct substitution
like this?

508
00:34:13 --> 00:34:19
You say, oh,
I need z prime.

509
00:34:15 --> 00:34:21
z prime equals,
well, I better use the quotient

510
00:34:18 --> 00:34:24
rule for differentiating that.
And, it comes out this long,

511
00:34:22 --> 00:34:28
and then either a long pause,
what do I do now?

512
00:34:26 --> 00:34:32
Because it's not at all obvious
what to do at that point.

513
00:34:30 --> 00:34:36
Or, much worse,
two pages of frantic

514
00:34:32 --> 00:34:38
calculations,
and giving up in total despair.

515
00:34:37 --> 00:34:43
Now, the reason for that is
because you tried to do it

516
00:34:40 --> 00:34:46
making a direct substitution.
All you have to do instead is

517
00:34:45 --> 00:34:51
use it, treat it as an inverse
substitution,

518
00:34:48 --> 00:34:54
write y equals zx.
What's the motivation for doing

519
00:34:51 --> 00:34:57
that?
It's clear from the equation.

520
00:34:54 --> 00:35:00
This goes through all of
mathematics.

521
00:34:57 --> 00:35:03
Whenever you have to change a
variable, excuse me,

522
00:35:00 --> 00:35:06
whenever you have to change a
variable, look at what you have

523
00:35:05 --> 00:35:11
to substitute for,
and focus your attention on

524
00:35:08 --> 00:35:14
that.
I need to know what y prime is.

525
00:35:12 --> 00:35:18
Okay, well, then I better know
what y is.

526
00:35:15 --> 00:35:21
If I know what y is,
do I know what y prime is?

527
00:35:19 --> 00:35:25
Oh, of course.
y prime is z prime x plus z

528
00:35:22 --> 00:35:28
times the derivative of this
factor, which is one.

529
00:35:26 --> 00:35:32
And now, I turned with that one

530
00:35:31 --> 00:35:37
stroke, the equation has now
become z prime x plus z is equal

531
00:35:36 --> 00:35:42
to F of z.
Well, I don't know.

532
00:35:40 --> 00:35:46
Can I solve that?
Sure.

533
00:35:42 --> 00:35:48
That can be solved because this
is x times dz / dx.

534
00:35:48 --> 00:35:54
Just put the z on the other
side, it's F of z minus z.

535
00:35:53 --> 00:35:59
And now, this side is just a

536
00:35:55 --> 00:36:01
function of z.
Separate variables.

537
00:36:00 --> 00:36:06
And, the only thing to watch
out for is, at the end,

538
00:36:03 --> 00:36:09
the z was your business.
You've got to put the answer

539
00:36:07 --> 00:36:13
back in terms of z and y.
Okay, let's work an example of

540
00:36:11 --> 00:36:17
this.
Since I haven't done any

541
00:36:13 --> 00:36:19
modeling yet this period,
let's make a little model,

542
00:36:17 --> 00:36:23
differential equations model.
It's a physical situation,

543
00:36:21 --> 00:36:27
which will be solved by an
equation.

544
00:36:24 --> 00:36:30
And, guess what?
The equation will turn out to

545
00:36:27 --> 00:36:33
be homogeneous.
Okay, so the situation is as

546
00:36:32 --> 00:36:38
follows.
We are in the Caribbean

547
00:36:34 --> 00:36:40
somewhere, a little isolated
island somewhere with a little

548
00:36:39 --> 00:36:45
lighthouse on it at the origin,
and a beam of light shines from

549
00:36:44 --> 00:36:50
the lighthouse.
The beam of light can rotate

550
00:36:48 --> 00:36:54
the way the lighthouse beams.
But, this particular beam is

551
00:36:53 --> 00:36:59
being controlled by a guy in the
lighthouse who can aim it

552
00:36:57 --> 00:37:03
wherever he wants.
And, the reason he's interested

553
00:37:01 --> 00:37:07
in aiming it wherever he wants
is there's a drug boat here,

554
00:37:06 --> 00:37:12
[LAUGHTER] which has just been
caught in the beam of light.

555
00:37:13 --> 00:37:19
So, the drug boat,
which has just been caught in a

556
00:37:16 --> 00:37:22
beam of light,
and feels it'd a better escape.

557
00:37:20 --> 00:37:26
Now, the lighthouse keeper
wants to keep the drug boat;

558
00:37:24 --> 00:37:30
the light is shining on it so
that the U.S.

559
00:37:27 --> 00:37:33
Coast Guard helicopters can
zoom over it and do whatever

560
00:37:31 --> 00:37:37
they do to drug boats,
--

561
00:37:34 --> 00:37:40
-- I don't know.
So, the drug boat immediately

562
00:37:37 --> 00:37:43
has to follow an escape
strategy.

563
00:37:39 --> 00:37:45
And, the only one that occurs
to him is, well,

564
00:37:42 --> 00:37:48
he wants to go further away,
of course, from the lighthouse.

565
00:37:47 --> 00:37:53
On the other hand,
it doesn't seem sensible to do

566
00:37:50 --> 00:37:56
it in a straight line because
the beam will keep shining on

567
00:37:54 --> 00:38:00
him.
So, he fixes the boat at some

568
00:37:57 --> 00:38:03
angle, let's say,
and goes off so that the angle

569
00:38:00 --> 00:38:06
stays 45 degrees.
So, it goes so that the angle

570
00:38:05 --> 00:38:11
between the beam and maybe,
draw the beam a little less

571
00:38:11 --> 00:38:17
like a 45 degree angle.
So, the angle between the beam

572
00:38:16 --> 00:38:22
and the boat,
the boat's path is always 45

573

574

575
00:38:20 --> 00:38:26
degrees, goes at a constant 45
degree angle to the beam,

576
00:38:26 --> 00:38:32
hoping thereby to escape.
On the other hand,

577
00:38:30 --> 00:38:36
of course, the lighthouse guy
keeps the beam always on the

578
00:38:36 --> 00:38:42
boat.
So, it's not clear it's a good

579
00:38:40 --> 00:38:46
strategy, but this is a
differential equations class.

580
00:38:44 --> 00:38:50
The question is,
what's the path of the boat?

581
00:38:48 --> 00:38:54
What's the boat's path?
Now, an obvious question is,

582
00:38:52 --> 00:38:58
why is this a problem in
differential equations at all?

583
00:38:57 --> 00:39:03
In other words,
looking at this,

584
00:38:59 --> 00:39:05
you might scratch your head and
try to think of different ways

585
00:39:04 --> 00:39:10
to solve it.
But, what suggests that it's

586
00:39:09 --> 00:39:15
going to be a problem in
differential equations?

587
00:39:13 --> 00:39:19
The answer is,
you're looking for a path.

588
00:39:17 --> 00:39:23
The answer is going to be a
curve.

589
00:39:20 --> 00:39:26
A curve means a function.
We are looking for an unknown

590
00:39:24 --> 00:39:30
function, in other words.
And, what type of information

591
00:39:29 --> 00:39:35
do we have about the function?
The only information we have

592
00:39:34 --> 00:39:40
about the function is something
about its slope,

593
00:39:38 --> 00:39:44
that its slope makes a constant
45° angle with the lighthouse

594
00:39:44 --> 00:39:50
beam.
Its slope makes a constant

595
00:39:53 --> 00:39:59
known angle to a known angle.
Well, if you are trying to find

596
00:40:04 --> 00:40:10
a function, and all you know is
something about its slope,

597
00:40:09 --> 00:40:15
that is a problem in
differential equations.

598
00:40:13 --> 00:40:19
Well, let's try to solve it.
Well, let's see.

599
00:40:16 --> 00:40:22
Well, let me draw just a little
bit.

600
00:40:19 --> 00:40:25
So, here's the horizontal.
Let's introduce the

601
00:40:23 --> 00:40:29
coordinates.
In other words,

602
00:40:25 --> 00:40:31
there's the horizontal and
here's the boat to indicate

603
00:40:30 --> 00:40:36
where I am with respect to the
picture.

604
00:40:35 --> 00:40:41
So, here's the boat.
Here's the beam,

605
00:40:38 --> 00:40:44
and the path of the boat is
going to make a 45° angle with

606
00:40:44 --> 00:40:50
it.
So, this is the path that we

607
00:40:47 --> 00:40:53
are talking about.
And now, let's label what I

608
00:40:51 --> 00:40:57
know.
Well, this angle is 45°.

609
00:40:54 --> 00:41:00
This angle, I don't know,
but of course I can calculate

610
00:41:00 --> 00:41:06
it easily enough because it has
to do with, if I know the

611
00:41:05 --> 00:41:11
coordinates of this point,
(x, y), then of course that

612
00:41:11 --> 00:41:17
horizontal angle,
I know the slope of this line,

613
00:41:15 --> 00:41:21
and that angle will be related
to the slope.

614
00:41:22 --> 00:41:28
So, let's call this alpha.
And now, what I want to know is

615
00:41:29 --> 00:41:35
what the slope of the whole path
is.

616
00:41:35 --> 00:41:41
So, y prime-- let's call y
equals y of x,

617
00:41:42 --> 00:41:48
the unknown function whose
path, whose graph is going to be

618
00:41:50 --> 00:41:56
the boat's path,
unknown graph.

619
00:41:54 --> 00:42:00
What's its slope?
Well, its slope is the tangent

620
00:42:00 --> 00:42:06
of the sum of these two angles,
alpha plus 45°.

621
00:42:08 --> 00:42:14
Now, what do I know?
Well, I know that the tangent

622
00:42:11 --> 00:42:17
of alpha is how much?
That's y over x.

623
00:42:14 --> 00:42:20
In other words,

624
00:42:16 --> 00:42:22
if this was the point,
x over y, this is the angle it

625
00:42:19 --> 00:42:25
makes with a horizontal,
if you think of it over here.

626
00:42:23 --> 00:42:29
So, this angle is the same as
that one, and it's y over,

627
00:42:26 --> 00:42:32
its slope of that line is y
over x.

628
00:42:30 --> 00:42:36
So, the tangent of the angle is
y over x.

629
00:42:32 --> 00:42:38
How about the tangent of 45°?
That's one, and there's a

630
00:42:36 --> 00:42:42
formula.
This is the hard part.

631
00:42:38 --> 00:42:44
All you have to know is that
the formula exists,

632
00:42:41 --> 00:42:47
and then you will look it up if
you have forgotten it,

633
00:42:44 --> 00:42:50
relating the tangent or giving
you the tangent of the sum of

634
00:42:48 --> 00:42:54
two angles, and you can,
if you like,

635
00:42:50 --> 00:42:56
clever, derive it from the
formula for the sign and cosine

636
00:42:54 --> 00:43:00
of the sum of two angles.
But, one peak is worth a

637
00:42:57 --> 00:43:03
thousand finesses.
So, it is the tangent of alpha

638
00:43:02 --> 00:43:08
plus the tangent of 45°.
Let me read it out in all its

639
00:43:06 --> 00:43:12
gory details,
divided by one,

640
00:43:08 --> 00:43:14
so you'll at least learn the
formula, one minus tangent alpha

641
00:43:12 --> 00:43:18
times tangent 45°.

642
00:43:15 --> 00:43:21
This would work for the tangent
of the sum of any two angles.

643
00:43:20 --> 00:43:26
That's the formula.
So, what do I get then?

644
00:43:23 --> 00:43:29
y prime is equal to the tangent
of alpha, which is y over x,

645
00:43:27 --> 00:43:33
oh, I like that combination,
plus one, divided by (one minus

646
00:43:32 --> 00:43:38
y over x times one).

647
00:43:37 --> 00:43:43
Now, there is no reason for
doing anything to it,

648
00:43:40 --> 00:43:46
but let's make it look a little
prettier, and thereby,

649
00:43:44 --> 00:43:50
make it less obvious that it's
a homogeneous equation.

650
00:43:48 --> 00:43:54
If I multiply top and bottom by
x, it looks prettier.

651
00:43:52 --> 00:43:58
x plus y over x minus y equals
y prime.

652
00:43:55 --> 00:44:01
That's our
differential equation.

653
00:44:00 --> 00:44:06
But, notice,
that let step to make it look

654
00:44:02 --> 00:44:08
pretty has undone the good work.
It's fine if you immediately

655
00:44:06 --> 00:44:12
recognize this as being a
homogeneous equation because you

656
00:44:10 --> 00:44:16
can divide the top and bottom by
x.

657
00:44:12 --> 00:44:18
But here, it's a lot clearer
that it's a homogeneous equation

658
00:44:16 --> 00:44:22
because it's already been
written in the right form.

659
00:44:20 --> 00:44:26
Okay, let's solve it now,
since we know what to do.

660
00:44:23 --> 00:44:29
We're going to use as the new
variable, z equals y over x.

661
00:44:27 --> 00:44:33
And, as I wrote up there for y

662
00:44:33 --> 00:44:39
prime, we'll substitute z prime
x plus z.

663
00:44:40 --> 00:44:46
And, with that,
let's solve.

664
00:44:44 --> 00:44:50
Let's solve it.
The equation becomes z prime x

665
00:44:50 --> 00:44:56
plus z is equal to z plus one
over one minus z.

666
00:44:57 --> 00:45:03
We want to separate variables,

667
00:45:04 --> 00:45:10
so you have to put all the z's
on one side.

668
00:45:07 --> 00:45:13
So, this is going to be x,
dz / dx equals this thing minus

669
00:45:11 --> 00:45:17
z, which is (z plus one) over
(one minus z) minus z.

670
00:45:14 --> 00:45:20
And now, as you realize,

671
00:45:18 --> 00:45:24
putting it on the other side,
I'm going to have to turn it

672
00:45:22 --> 00:45:28
upside down.
Just as before,

673
00:45:24 --> 00:45:30
if you have to turn something
upside down, it's better to

674
00:45:28 --> 00:45:34
combine the terms,
and make it one tiny little

675
00:45:31 --> 00:45:37
fraction.
Otherwise, you are in for quite

676
00:45:35 --> 00:45:41
a lot of mess if you don't do
this nicely.

677
00:45:39 --> 00:45:45
So, z plus one minus z,
that gets rid of the z's.

678
00:45:43 --> 00:45:49
The numerator is one minus z
squared over one minus z,

679
00:45:48 --> 00:45:54
I hope, one,
is that right,

680
00:45:50 --> 00:45:56
(one plus z squared) over (one
minus z).

681
00:45:56 --> 00:46:02
And so, the question is dz,

682
00:45:58 --> 00:46:04
and put this on the other side
and turn it upside down.

683
00:46:05 --> 00:46:11
So, that will be (one minus z)
over (one plus z squared) on the

684
00:46:10 --> 00:46:16
left-hand side and on the
right-hand side,

685
00:46:14 --> 00:46:20
dx over x.
Well, that's ready to be

686
00:46:17 --> 00:46:23
integrated just as it stands.
The right-hand side integrates

687
00:46:23 --> 00:46:29
to be log x.
The left-hand side is the sum

688
00:46:26 --> 00:46:32
of two terms.
The integral of one over one

689
00:46:30 --> 00:46:36
plus z squared is the arc
tangent of z,

690
00:46:34 --> 00:46:40
maybe?
The derivative of this is one

691
00:46:37 --> 00:46:43
over one plus z squared.

692
00:46:40 --> 00:46:46
How about the term z over one
plus z squared?

693
00:46:44 --> 00:46:50
Well, that integrates to be a

694
00:46:46 --> 00:46:52
logarithm.
It is more or less the

695
00:46:48 --> 00:46:54
logarithm of one plus 
z squared.

696
00:46:51 --> 00:46:57
If I differentiate this,
I get one over one plus z^2

697
00:46:54 --> 00:47:00
times 2z,
but I wish I had negative z

698
00:46:58 --> 00:47:04
there instead.
Therefore, I should put a minus

699
00:47:01 --> 00:47:07
sign, and I should multiply that
by half to make it come out

700
00:47:05 --> 00:47:11
right.
And, this is log x on the right

701
00:47:09 --> 00:47:15
hand side plus,
put in that arbitrary constant.

702
00:47:13 --> 00:47:19
And now what?
Well, let's now fool around

703
00:47:16 --> 00:47:22
with it a little bit.
The arc tangent,

704
00:47:19 --> 00:47:25
I'm going to simultaneously,
no, two steps.

705
00:47:22 --> 00:47:28
I have to remember your
innocence, although probably a

706
00:47:26 --> 00:47:32
lot of you are better
calculators than I am.

707
00:47:31 --> 00:47:37
I'm going to change this,
use as many laws of logarithms

708
00:47:35 --> 00:47:41
as possible.
I'm going to put this in the

709
00:47:38 --> 00:47:44
exponent, and put this on the
other side.

710
00:47:41 --> 00:47:47
That's going to turn it into
the log of the square root of

711
00:47:45 --> 00:47:51
one plus z squared.

712
00:47:48 --> 00:47:54
And, this is going to be plus
the log of x plus c.

713
00:47:53 --> 00:47:59
And, now I'm going to make,

714
00:47:55 --> 00:48:01
go back and remember that z
equals y over x.

715
00:48:00 --> 00:48:06
So, this becomes the arc
tangent of y over x equals.

716
00:48:04 --> 00:48:10
Now, I combine the logarithms.

717
00:48:09 --> 00:48:15
This is the log of x times this
square root, right,

718
00:48:12 --> 00:48:18
make one logarithm out of it,
and then put z equals y over z.

719
00:48:16 --> 00:48:22
And, you see that if you do

720
00:48:19 --> 00:48:25
that, it'll be the log of x
times the square root of one

721
00:48:23 --> 00:48:29
plus (y over x) squared,

722
00:48:27 --> 00:48:33
and what is that?
Well, if I put this over x

723
00:48:30 --> 00:48:36
squared and take it out,
it cancels that.

724
00:48:34 --> 00:48:40
And, what you are left with is
the log of the square root of x

725
00:48:38 --> 00:48:44
squared plus y squared plus a
constant.

726
00:48:42 --> 00:48:48
Now, technically,

727
00:48:43 --> 00:48:49
you have solved the equation,
but not morally because,

728
00:48:47 --> 00:48:53
I mean, my God,
what a mess!

729
00:48:49 --> 00:48:55
Incredible path.
It tells me absolutely nothing.

730
00:48:52 --> 00:48:58
Wow, what is the screaming?
Change me to polar coordinates.

731
00:48:56 --> 00:49:02
What's the arc tangent of y
over x?

732
00:49:00 --> 00:49:06
Theta.
In polar coordinates it's

733
00:49:02 --> 00:49:08
theta.
This is r.

734
00:49:04 --> 00:49:10
So, the curve is theta equals
the log of r plus a constant.

735
00:49:09 --> 00:49:15
And, I can make even that
little better if I exponentiate

736
00:49:14 --> 00:49:20
everything, exponentiate both
sides, combine this in the usual

737
00:49:19 --> 00:49:25
way, the and what you get is
that r is equal to some other

738
00:49:24 --> 00:49:30
constant times e to the theta.

739
00:49:30 --> 00:49:36
That's the curve.
It's called an exponential

740
00:49:33 --> 00:49:39
spiral, and that's what our
little boat goes in.

741
00:49:37 --> 00:49:43
And notice, probably if I had
set up the problem in polar

742
00:49:42 --> 00:49:48
coordinates from the beginning,
nobody would have been able to

743
00:49:48 --> 00:49:54
solve it.
But, anyone who did would have

744
00:49:51 --> 00:49:57
gotten that answer immediately.
Thanks.