Page 139 - Schaum's Outline of Theory and Problems of Advanced Calculus
P. 139

130                            PARTIAL DERIVATIVES                         [CHAP. 6


                                                       !
                                                                        3
                                                            2
                                                                   2
                                                                2
                                   2
                                          @z
                                  @ z   @       @   x 3    ðx þ y Þð3x Þ ðx Þð2xÞ  2   3   1   2
                                                   2
                                 @x @y  ¼  @x @y  ¼  @x x þ y 2  ¼  2  2 2   ¼   2 2   ¼ 1at ð1; 1Þ
                                                                ðx þ y Þ
                              The result can be written z xy ð1; 1Þ¼ 1:
                              Note:Inthis calculation we are using the fact that z xy is continuous at ð1; 1Þ (see remark at the end of
                          Problem 6.9).
                     6.13. If f ðx; yÞ is defined in a region r and if f xy and f yx exist and are continuous at a point of r, prove
                          that f xy ¼ f yx at this point.
                              Let ðx 0 ; y 0 Þ be the point of r.  Consider
                                           G ¼ f ðx 0 þ h; y 0 þ kÞ  f ðx 0 ; y 0 þ kÞ  f ðx 0 þ h; y 0 Þþ f ðx 0 ; y 0 Þ
                              Define    (1)   ðx; yÞ¼ f ðx þ h; yÞ  f ðx; yÞ  (2)   ðx; yÞ¼ f ðx; y þ kÞ  f ðx; yÞ
                          Then         (3)  G ¼  ðx 0 ; y 0 þ kÞ   ðx 0 ; y 0 Þ  (4)  G ¼  ðx 0 þ h; y 0 Þ   ðx 0 ; y 0 Þ
                              Applying the mean value theorem for functions of one variable (see Page 72) to (3) and (4), we have
                                (5)  G ¼ k  y ðx 0 ; y 0 þ   1 kÞ¼ kf f y ðx 0 þ h; y 0 þ   1 kÞ  f y ðx 0 ; y 0 þ   1 kÞg  0 <  1 < 1
                                (6)  G ¼ h  x ðx 0 þ   2 h; y 0 Þ¼ hf f x ðx 0 þ   2 h; y 0 þ kÞ  f x ðx 0 þ   2 h; y 0 Þg  0 <  2 < 1
                              Applying the mean value theorem again to (5) and (6), we have
                                         (7)  G ¼ hk f yx ðx 0 þ   3 h; y 0 þ   1 kÞ  0 <  1 < 1; 0 <  3 < 1
                                         (8)  G ¼ hk f xy ðx 0 þ   2 h; y 0 þ   4 kÞ  0 <  2 < 1; 0 <  4 < 1
                              From (7) and (8)we have

                                              ð9Þ f yx ðx 0 þ   3 h; y 0 þ   1 kÞ¼ f xy ðx 0 þ   2 h; y 0 þ   4 kÞ
                              Letting h ! 0 and k ! 0in(9)we have, since f xy and f yx are assumed continuous at ðx 0 ; y 0 Þ,

                                                         f yx ðx 0 ; y 0 Þ¼ f xy ðx 0 ; y 0 Þ
                          as required.  For example where this fails to hold, see Problem 6.41.

                     DIFFERENTIALS

                     6.14. Let f ðx; yÞ have continuous first partial derivatives in a region r of the xy plane.  Prove that
                                        f ¼ f ðx þ  x; y þ  yÞ  f ðx; yÞ¼ f x  x þ f y  y þ   1  x þ   2  y
                          where   1 and   2 approach zero as  x and  y approach zero.
                              Applying the mean value theorem for functions of one variable (see Page 72), we have

                                   ð1Þ   f ¼f f ðx þ  x; y þ  yÞ  f ðx; y þ  yÞg þ f f ðx; y þ  yÞ  f ðx; yÞg
                                                                               0 <  1 < 1; 0 <  2 < 1
                                         ¼  xf x ðx þ   1  x; y þ  yÞþ  yf y ðx; y þ   2  yÞ
                          Since, by hypothesis, f x and f y are continuous, it follows that
                                        f x ðx þ   1  x; y þ  yÞ¼ f x ðx; yÞþ   1 ;  f y ðx; y þ   2  yÞ¼ f y ðx; yÞþ   2
                          where   1 ! 0,   2 ! 0as  x ! 0 and  y ! 0.
                              Thus,   f ¼ f x  x þ f y  y þ   1  x þ   2  y as required.
                              Defining  x ¼ dx;  y ¼ dy,we have  f ¼ f x dx þ f y dy þ   1 dx þ   2 dy:
                              We call  df ¼ f x dx þ f y dy the differential of f (or z)orthe principal part of  f (or  z).

                                        2
                     6.15. If z ¼ f ðx; yÞ¼ x y   3y, find  (a)  z;  ðbÞ dz: ðcÞ Determine  z and dz if x ¼ 4, y ¼ 3,
                           x ¼ 0:01,  y ¼ 0:02.  (d)How might you determine f ð5:12; 6:85Þ without direct computa-
                          tion?
   134   135   136   137   138   139   140   141   142   143   144