NUMERICAL METHODS- Theoretical & py, ct tog erative method is known as 14. The it (a) direct method (b) indirect method (c) none of these 15, The solution of a system of equations is obtained y cagoessive approximation method is known as (a) direct method (®) indirect method (©) both (a) and ©) (a) none of these Answers ld 2a 3b 4a 5b Ga Ta 8d lla 12a 13.a 14-b 15-b Bg | NUMERIcay s b gntroduetion. op applied mathematics and en gine Bom of finding one or more roots of pe SHERI face ; | pe prl | equation | | Q m of the real f(x)=0 see fe) m general, a nonlinear fons 1¢ x. But in most cases, itis aoa Sf the equation (1) and, therefore wee ott ait of (1) aumercially with any ne sed ork securacy- ThE numerical methods of finding ied degree of | Shed iterative methods. these roots are he function /(3)mmay have any one ofthe folloning fo : 3g forms: jy f(z) is an algebraic or polynomial E ay, 60 that function of degree n, fe) 20 40,04 a,x" +40, yx 40,2" n)are constants, real or complex, and where a,(i = 0,1 x? Txt], x2 425-4243 ete. are | 0,20. For exampl aiebraic functions. In such cases, th 205 called algebraic equation. 1 the equation ffs) =O%s Gi) /(x) isa transcendental function, ie. /(=) i a if) ental function, ie. f(x) isa function f(x) =a, + ax tage? +---t0 | are constants, real or complex, and all | | i = 0,1,2, | @ rexample, sinx+2x7-1, e* +logs+5 cle hascendental fanvtions. Here the equation f(s) =0 called Jn the function /(+) is 22 of the equation (D- | methods 10 ‘compute the or transcendental uation f(x) = 0if i a real root of ax no multiple reel, Ja) =0, f(a) 4 9- 1) root of (1) by ts, in gene a f(s) f(a) 20 then Dotert nds to be di wing (Wo Steps (i) Isolating the roots, i.e. interval [a, 6] containing one am (ii) Improving the vi refinding them to t nt the first step, we use finding the smallest possi only one root of the fo To imple of a continuous: function ‘Thoorem 1. If real values b} and /(a),/(0) are of opposite Gne real root of /(x) = 0 in (@, 6) a function /(x) is continuous in ‘signs, then there is at least Processes. . ration once {x,}ofiterates of a root « of the 6.2. It equation Lot the f(x) =O is produced by a given method. Then the error tho n th iteration is given by involved at @ ges 5.0 as n= a, we say that the iteration conve and the sequence {x, } converges to a. Otherwise, iteration is dives {the method of computation fai primary task is to find the condition of converdence If cy expressed Thus ov of the iteration processes. The error G1 can torms of Gy» Gat which we call error equation: If wo define h, by po 28 Ena, thom. ft ition of ty if Sar Mpproximates a, Hf the, iteration in tor? por bound for cs, then we can find an up} ‘his is called estimation of error. | @ order of of convergence and q ig k is known { Hie error constant. The iter ae rative method with p>1 depends of root. If this initial factory the iterative method dass wat ive met ok for the new computation | ative method is self correct, ie. if i ‘dnror in the calculations of iteration, thecmancs now initial approximation leading ta correct 43. Bisection Method. ‘A. Basic principle and formula method of bisection is-the most simplest iterative sethed. It is also known as half-interual ot Bolzano method. ths method is based on ‘Theorem 1 on the change of sign. Inthis method, we firat find out a sulficionty small interval Jydp]eontaining the required root « of the equation (1). Then f(ay)f(by) <0 and /'(x) has the same sign in lobJand so f(x) is strictly monotonic in [a0.b) ‘To genorate the sequence {%,} of iterates, "° put 0 by and 1 = F(ao + bo)and find flay)» UE Pls) and 'ls))are of opposite signs, then set % = ay b= it fle)and Il f h]=[ag,x,]. On the other hand, rite signs then put a =* ae see that [aj,b,]contains the =aNUMERICAL METHOD Theoretical py, 1 | Next set *2= (01+ br)and repeat the above proce, al we obtain 1 net = 3 (On + Pn) veyed until the interval = ‘i ing t with desired accuracy with x, > a as n— co ane we des ae B. Convergence of bisection method Suppose the interval [a, £(@n)f(bx) <0. contains the root agg 1 egent Let nat = 5(s +b). | omer ee yantage. The method - ‘ey of iteration to obtain moderately accurate If f(@y)F(%n41) <0, then set Gq.) = Ons On.) 7 ae Pon aEoicaes ioderately accurate results and On the other hand, if /(xp41)/(bq) <0, then Exompte... Find the root of th xtan=128, ere =) lies in the interval (0,1), correct to four places of decimal Inet = Fra ePoad using bisection method. WB.UT, M(CS)-312, 2005, ‘Thus in any case M(CS)-301, 2015, BCS)-401, 2015) ae lOn bath Manadll Solution. Let f(x) = xtanx-1-28 and = (0) = -1-28 < 0, f(1) = 0-277408>0 1 Brn ~ Anat <7 (Pn ~ An) < So a root lies between 0 and 1. 1 Take a)=0, by=1 so that x gle+t)=08 Since 1(05) = -1.006849 < Oand 0) > 0, the root lies between 0.6 then If Epsibe the error in approximating a by <, ast Sp Oe < PSM 7 0as n+@, ‘Thus the iteration converges. sat) thas we have #2 = (05 1) jl nt Proceeding in this way, wre obtain the falling table “E = g so the convergence in bisection methods neat. | 0, +b, fen) G, 32 * ©. Geometrical Interpretation of Bisection method fla,) <0 MaKau.r, mos-401, 21 5 ; i am | L000 5 81308 Let f(2)=0 be the equation and flap) f(tp)<0- THE | 06 1 a1 root lies between ap and b, as shown in the figure. Tak O78 1 ots ea y+ a 6 |-O a= ott, we divide the interval [ag, bp] into two hilt Gap i comet 0.129819 to locate a new sub-interval containing the root. 7 0.9375 |_4an UMERICAL METHOD Theoretical & Prog, q 5 0.9375 | 0.96875 6 dos7s | 0.961675 | 0.945312 | 0.028505 7 .9376 | 0.915312 | 0.941406 | 0.012819 8 o.9375 | 0941406} 0.939468 | 0.004855 9 o.9a7s | 0.939453 | 0.938477 | 0.000893 io | 09875 | 0.998477 | 0.937988 | -0.00108 11 | 0.937988] 0.998477 | 0.988292 | ~0.000006 ye | 0.998292] 0.938477] 0.938954 | 0.000398 13 | 0.938292] 0.938954] 0.938293 | 0.000151 11 } 0.938282] 0.938293} 0.938263 | 0.000028 Thus a real root of the given equation is 0.9388 correct ty four decimal places. 6.4 Regula-Falsi Method. A. Basic principle and formula ‘The regula-falsi method or false position method is also (WB.U.T., CS-312,2003) sometimes referred to as the method of linear ” interpolation and itis the oldest method for ‘computing real roots of an equation f(x) =0. To find a real root « of (3) =0, we first choose ao sufficiently small interval [0,3)]in which the root (2S) «a lies. Then /(x)and_/(x,) must be of opposite signs so that ‘S(%0)f(%1)<0and the graph of f(x) must cross the x-axis between x=xand x=x,. Since the interval [x9,%]'8 sufficiently small, the portion of the curve between 4[%0,f()) and 2[x,,/(x))] can be approximated by a secant line (straight line) and so the intersection of the secant AB with the »-4%i gives an approximate value x, ,say, of the root. (70. £0) 20.0 =p, We derive putting (s1=x0) (5) ) = 0, then isa 00tof /0) <0;ethere (s) 0 sf ft)? A(x.) and F(x2) are of opposite signs the root «between % On the other hand iff(x))and f(x) aro of Me and xp and in this ease we set xy =x, ase eal jgns, the root lies apposite SIEM, between. and x» and thus, in ither case, . flss)fle) <0 Honce the next approximation of the root, say xq lies ietween.xy and xp and get (x2) Fe) fey) “0 ‘The general formula based on the above process is wx, - fl) fg, en yynehaee Baek 0 FS fee) 8 @ This is regula falsi iteration formula. The processis repeated ntl the root is obtained to required degree of accuracy. B. Convergence of regula falsi method. let @ be a simple root of the equation /(x)=0. Then Matting, =a+e, in (7), we get Enel =n Fg ce,)- Mat ent)\ | « nding Fin Paylor's: 1 al (0) y Pal () Te) Ney He) =e, vets =A Le), 27) so that: Ener = C Ena re Co BL) where C= 5 Cy The ition (9) is cal =m=4(14¥5) Neglecting the minus sign, convergence of {x,} is m= 1-618, Also from (10) we get Azone and we have negleet d the error equa To find the order of convergence, os geal) | 19 Fart e|-[euaren (19) , we find that the order of the acto“ or pr! © equatio 3 =2s-5coby tgives the following table 0944 isa root of the ices, given equation corres No, of teratio 0] fof un @ | 1 16 | 2.05882 | -0.99084 2 | 205882] 3 | -0.39084] 16 | 2.08126 |-o..47244 3 [208126 | 3 |-o.ra7244] 16 | 208064 | -0 054007 4 | 208961 ] 3 |-0.051007] 16 | 2.09274 | -oo201s 5 |2.00274 | 3 |-o.oz0198} 16 | 2.09988 |-0.007401 6 | 209388 | 3 |-0.007491} 16 | 2.09190 |-0.002806 7 | 2.09130} 3 |-o.002600} 16 | 2.09145 |-0.00118 8 {209145 | 3 |-o.ooriss] 16 [2.094851 1 upto fourNUMERICAL METHOD Theoretical & py, 200 ay 6.5. Newton-Raphson method. principteand formula, (WB.U-T:, CS.319 9 ‘an initial approximation of the desired yooy dx, = Xp +h is the correct rogy* ot A. Basic. Let robe the equation /(x) = 0am flea) =0 Loa fey +h) = 0,(min(x1-¥0) << max(ar,xo)) 2 ie, fleo)*W0)* Sp “=0 (0) + by Taylor's series, expansion Neglecting the second and higher order terms, we obtain f (x0) +hf'(%0) = 0» f(xo) ie, h=- 7'(0) ‘Thus a better approximation of the root « is _ fo) 0" Fag) ay Repeating the above process and replacing x, by x,we obtain the second approximation of the root as f(x) Fe)” Proceeding in this way, we get the successive approximations ty =x) - 3554," Xng) Where Aen) wn 2) Tn)" Fast = Fn — provided /"(x,,) # 0,n = 1,2; The result (11) is known as Newlon-Raphson iteration formula. If [a,b] be the initial interval in which the ° given equation /(x)=0 lies and /'(x) #0, then the i approximation may be started with x9 = @ or 6- root, cof the initial netrical meaning of Newton p Geom ‘ah re snecurve ¥=f(*)cuts the x axig at _ formula ‘a root of the equation (x)= p° *M® Point x pat AN, Me) =0, = a0 Al Se) eo 0 @ Sut hy x the tangent to the curve at the point P,,f(z,))euts xis at the point * =*,,1 and isinclined at an angle owith a ee en fen) +(x,) = tand =e i ala athat Alea) ne re)" ‘Accordingly, the Newton-Raphson method may also be called the method of tangents. C. Convergence of Newton-Raphson method. [W.B.U.T., CS-312,2003,2007, ‘M(CS)-401,2013,2014,2015) Let «be a root of the equation f(x)=0 sHla)=0 ten flsn +4-Hn) =O 2 ie, fan) + (a —¥n)f"(%n) esl rtes) =o ‘Taylor's theorem [info x,|-<&,, < masa.) by a3)202 NUMERICAL METHOD -Theoretical & py, OO We have, therefore, from (12) and (13), e mua = Hens) Putting a—x, = &, anda@~ Xp. = Enet, We get (uy which is the error equation. If the iteration converges then x ener| 1 /°(@) mre: |" 2 7(@) a En PU ASN 050thap Hence Newton-Raphson method is a second order iteration process. So the convergence is quadratic and the constan, 1 f(a) asymptotic, error is equal to 3 7(q) D. Advantage and disadvantage of Newton-Raphsor method. AKA.U.T., MCS-401,2014 ‘Advantage. The rate of convergence of this method is quadratic. So the method converges more rapidly than other numerical method. Disadvantage. In this method, the initial approximation must be chosen very close to the root; otherwise, the method will fail. Since the method depends on the derivative /“(x), i may not be suitable for 2 function /(x) whose derivative is difficult to compute. Also the method fails if /‘(x) = 0 or smal in the neighbourhood of the root. Note. When '(x) is large in the neighbourhood of the re root, ie., when the graph of function y = f(x) is nearly vertical then it crosses the x-axis. In this case, the method is very wseft! and the correct value of the root can be obtained more rapidls Example.3. Find the smallest positive root of the equatic? ‘ax? — 9x2 + 8 = 0 correct to four places decimal,using New’ Raphson method. [W.B.U.T., CS-312,2009 0) = 0 yow (0) “4. Si ewe and 2. Choose x, =1 © A Positive root lies From () Gx? 9x? =f 1° oxy — 18% o(1-22222)" ~ Hence positive real root of the given equation correct to four decimal places is 1.2261. E. Newton-Raphson method for finding an assigned root ofa positive real number. [W.B.ULT,, CS-312,2009] root of a real number R. Suppose we are to find the Solet x= VR204 ‘The Newton-Raphson iteration formula (13) gives am-R (ty -xj|<€, ebeing the desired degree of accuragy with [x41 It is obvious from (17) that for finding the square ro, any positive number R (where m= 2), we have Bnet = B[%e ty fp MOL Ot of Example.4. Find the cube root of 10 upto 5 significant figures by Newton-Rapshon method. Solution. Let x=¥/10 x9-10=0 Let f(x) =x*-10. f(x) = 3x" Now f(0) = -10, f(1) = -9, f(2) = ~2, /(8) = 1750 that a real root of f(x) = O lies between 2 and 3. Hence using the iteration formula (16) with m = 3, we have 2x3 +10 i Spay = AN, n= 01,2, 1 axe i} which gives with Xo = 2 ox8 x = 20410 _ 916666 Bxz 22, * Similarly, x) = 4216666) +10 = 215450. 3(2.16666) xy = 215443, x, = 215443 Hence we have x=4/10=21544 correct to five significa" figures. Se sotto” Let x= geir=0 pot f= 17 _ pe=8e y= 17, FC) ==16,2)=-9, (210 il goa root lies between 2 and 3, We choose *) =2 i From, NB iterative formula, Xn = n=0,1,2, f we got 2x} 417 Se = a7 53 Qa +17 _ 22°17 _ ong Lox, = PANT Bee ae 8x2 8x2! 2x} +17 2.78) +17 _ 95505 xy = HA 2 axe 3(2.75)" 817 ay = 228820417 9518 (2.5826) = 25t 13)°+17 _ 95713 492.5713)" So the root of x? - 17=0 is 25718 Hence the cube root of 17 is 25719Ve 206 NUMERICAL METHOD -Theoreticat « Iutusrrative EXaMPLes Ex.1. Find a real root of the transcendental x* + 2x~2=0, correct upto two decimal places usin, method. Solution. Let f(x) = x* +2x-2. ‘Then f(0-5) = -0-293,f(1) =1 So a reul root lies between 0.5 and 1. equa ti f bisegi ” tion 0-541 Take ey = 0-5,b) = 1 so thet 1 = = 0-75, Again, since f(0-75) = 0-3059, so the root lies hetyy, and 1, Proceeding in this way, we construct the foliowi, een 0.75 Ing table | saredo, ony rn) | (flan) <9) |(A)>9) oO 05 1 0.75 0.3059 i 05 | 0.625 | ~0.0045 2 0.625 0.687 0.1466 3 | 0.625 0.687 0.656 0.0704 4 | 0.625 0.656 0.640 0.0325 5 0.625 | 0.640 i 0.632 0.0041 6 0.625 [0632] 0.628 0.0036 t 7 0.625 [0.628 | 0.626 ‘Then the root correct upto twe decimal places is 0.63. Ex. Find the smallest positive root of the equation e* = 4sinz correct to four decimal places by bisection method. Solution. The given covation is Asinx—e* =0- Let (x)= 4sinx—e*, vO SOLITON OP ALGEBRA FQuanio xe INS =-1, £(Q) = 064760, ince I Ws and 1. Take ay = (0.5) = 026898, 0 the root lies between 0 and and 05, following tables (howe O58 0.25 0.5 0.25 0.375 0.33333 0.375 0.35416 0.375 0.36979 eer 0.36979 | 0.25 0.375 0.83333, 0.35416 0.36458 0.36979 0.37240 0.37109 0.37044 0.37077 0.37060 0.37052, 0.37056 ‘Thus the required smallest positive root is 0.9706 correct to four decimal places 0.29440 O.01010 0.08690 ) ~0.03770 ~0.01360 0.00168, 0.00420 0.00126 ~0.00240 0.00047 0.00011 ~0,00007 Exd. Find the rot ofthe equation Sx—cosx-1=0 thatlesbaween ing bisection method Sand 1 correct to four places ei Ta CS.301, 2009) Solution, Let f(x) =3x~cosx~1 + (0) = -2, f(2) = 14597NUMERICAL METHOD -Theoretica 208 ii So a root lies between 0 and 1 Take ay = 0, by =1 1 = =—(0+1)=05 m=Z0+) Now f(x) = -0.3776 So the root lies between’0.5 and 1 O5+1 ‘Thus we have % = 9 Proceeding in this way, we obtain the following table = 0.75. No of a, by iteratior| (f(a,)<0) | (/(2.) > 0) qu) 0 0 1 0.5 ~0.3776 1 0.5 1 0.75 0.5183 2 0.5 0.75 0.625 0.0640 3 0.5 0.025 0.5625 | -0.1584 4 0.5625 | 0.625 0.59375 | -0.0476 5 0.59375 | 0.625 0.60938 | 0.0081 6 0.59375 | 0.60938 | 0.60157 | -0.0198 7 0.60157 | 0.60938 | 0.60548 | -0.0058 8 0.60548 | 0.60938 | 0.60743 | -0.0012 9 0.60548 | 0.60743 | 0.60646 | -0.0023 10 | 0.60646 | 0.60743 | 0.60695 | -0.0005 11 0.60695 | 0.60743 | 0.60719 0.0003 12 0.60695 | 0.60719 0.60707 | -0.0001 13 0.60707 _| 0.60719 | 0.60713 _|_0.0004 four decimal ‘Thus the required root is 0.6071, correct upto places. Ex.4. Using Bisection method obtain a root betw [MA.K.A.U.T,.M(CS)-40- eon Land 2 the equation e* -3x=0. 1, 2016) pri0bt SOLUTIONS OF AL otha Puan: Ng jon. Let (x)= ~3x f(0)=1>0 (1) =-0.281718 <0, 0.5 0.5625 0.59375 0.609375 0.5625 0.59375 0.609375 06171875 0.621094 0.619141 0.619141} 0.618164 0.619141] 0.6186525 0.00282 0.00009, 0.0010262 0.618164 «. The root is 0.62 correctup to 2 decimal places. 4 Ine -2=0 byNewton-Raphson Ex5. Find a positive root of * nethod correct upto six significant figure. eto Se LAK A.UT M(CS) 4012015 2016 Solution. Let f(x) =x+Inx-2 fQ)=-1<0 f(2)= 0.693147 >0 So, a positive lies between 1&2 We choose xy =1.5 f(xyal+ 4NUMERICAL METHOD -Theoretical & p,, ty ton Raphson method fn) _ F'n) By New Xqet =n (xq lan = 2 ent tn 14%, eee 2 x, InXy + 2%n ~ 14%, By = 80 Nn 14% x, = So toln¥o = 1.556721 Q 1+X0 x _ 8x = 1.556714 Te Similarly, xy = 1.557146 xy = 1.557146 +: The required Ex.6. Find a root of the eq Newton-Raphson method corres root is 1.55715 correct up to 6 singnificant figure uation xsinx-+cosx=0 using .ct upto 5 places of decimal. [W.B.U.T.,CS-312, 2004 Solution. Let f(x) = xsinx + cos¥ f(a) = xcosx +sine sine = Fe phson iteration formula SX So the Newton-Ra En) y= 0,1, 2,0 n+l, = % 7 . F'(%n) gives . x, sin, +C08%, = in SIN y a Xavi = ¥n ~ %y CO8Xy 1 a) =X, -—— tanx, oe x | ERE yw "BRAIC py CE WATION. Now (0) = 1, FQ) = 138, /(2) <4 ft go a oot His between 2and , choos gen we have, from (1), “056 @ x0 = 25, for quick convergey, ce, 1 2 2.847022-——4_ = J 2 BaT0nD "2847022 i = 2.799175 gimilarly *5 = 2.798386 xy = 2.798386 Hence a positive real root of the gi ive decimal places is 2.79839. given equation correct to x7. Find out the root of the following equation using Regula falsi method x® — 5x — 7 = 0 that lies between 2 and Seorest to4 decimal places. [W.B.U-L.CS.312, 2006, 2009) Solution. Let f(x) = x°-5x-7 We choose xg = 2,21 =3 sothat f(x) = -9, f(x1) =5- Then the regula-falsi iteration f(%n) x, _— Ma) BSF Fe) Ment) formula fq — teal 2 0) ives 5 _(g_) 2612857 3-H a=tive, following table = the Noof ea x, iteration(n) | (én-al®)| (Asn) 1 2 eo 5 2 aeiss7 | 3 | -175874q 3 2735635 | 3 | 0.203504 5 | 4 zrscor2 | 3 | 0.022474) 5 5 azsiz08 | 3 | 0.002444) 5 | 6 zas7as2 | 3 | 0.001 5 Hence a real root of the given eq) decimal places is 2.7473. : Ex.8. Find the root of the equation Regular-False method. xe" =cosx using the (WB.U.T. MCS-401,2009, | Solution. ‘The given equation can be written as | cosx-xe* =0 Let f(x) = cosx - xe* «. f(0) = 1, (1) = -217798 Soa real root lies between 0 and 1 We choose x9 =0,x, =1 © fxo) = 1, f(x) = -217798 Then Regula-falsi iterative method fs) Tem) et 0) a) gives -217798, -217798-1 = 031467 (1-0) « fl%s) = 051987 now f2 go the F00t Ties Between 0.1467. , From (2), -217798 Se sg 7172217718 - 051957 "~ 081467) 2044673 again f(s) = 020356 go the root lies between 0 ‘then applying (1), we get xy = 049402 i 44673 and 1, Repeating the process, we can obtain 5 = 050995. x5 = 051520, x, - 051699 = 051767, xy = 0517 xy = 051748, ‘Thus the required root is 0.51 places. Ex. Find the root of the equation 3:- Regula-Falsi method. Correct upto four decimal 05x-1=0 using (W.B.U.T. Mcs.201, 2007, Solution. Let f(x) = 3x-cosx-1 - f(0) = ~2, (1) = 1.4597 Soa real root lies between 0 and 1 wechoose x, =0,x;=1. Then the iteration formula f(%n) Suet = Xn (en 3 ne rn A Feller eeu tives =1-__ 14597 2° 1-Ti597-(~ayt?) = 0.578085 “A(%2) = 0.103256 So the root lies between 0.578085 and 1.HOD Theoretical & py, NUMERICAL M a4 ‘thon again applying (D), we et __14597___(1 0.578085) 5 = 1 Ta (-0103256) | = 0.605959 and (ss) = 200408 ., From (I), __1.4597___(1 -.0,605959) | 24 =1> 74597 -(-0.00408) = 0.607057 « f(4) = 0.000159 _ 14597 __(q - 0.607057) Hence *5 = 1~ 77597 (0.000159) | = 0.607099 | Hence the roots of the given equation is 0.6071, correct upts four decimal places. Ex.10. Find the root of the equation xe"-3=0 that lies between Land 2, correct to 4 significant figure using the methed of False position. Solution. Let (x) = x2*-3 Here we choose xp =1,.xy =2 sothat f(x) =-02817, f(x,) = 11.7781 So the iteration formula for false position method Mea) _ H&n)-f(%n-a) Xnvt = Hn = (ta tnt) n= 12 gives 117781 11.7781 (-02817) x =2- (2-1) = 102336 1 o2a7 “isan 1.04247 2 |4 ie No a wee 1.04595 ° M.7vail obmd goin, 2 2 1.04 hus the required root correct tofoursigniicart figures 49 geil. Find the root of the equation ef =2+1 erect w 4 paces of decimal, using Newton-Raphson meth near x = [W.B.U-T., CS-312, 2005, Solution. The given equation is eF =2x+1 ie, ef -2x-1=0 Let f(x) =e -2x-1 2 f(s et -2 So the Newton-Raphson iteration formala fe) nao, f'(%,) _ eeeberd re) fe ne +1 _ o b on -2 ‘or quick convergence, we choose % = 3 Fao tivesCAL SOLITONS OF Aten Chan ations From (1), we get L5el* ~e 41 = Mel Set § 1305003 er soos sence OPEC to four signi m= T305003 — y Bae ee seciproal of mato, form 'mber N and ben = 1259059 Similarly x, = 125906 value of 33° Uhre vigniticant Figur Hence the root of the given equation near x=1ig 1y55 | solution. Ex.12. Using Newton - Raphson method, find the va eof ye fetes Ineo Solution. Let x=¥Yi2 ua x oxt=l2 | [et se)=t-" So let f(x)=x'-12 1 2s @)=a? (= Then the Newton ; oe iteration formula | Sothe Newton - Raphson iteration formula Xn, Xai = ¥n Fy ane 2 1 Fe,) 2 = Oh? gives x,'-12 4x,° _ Sxf +12 “5 a =x, +x,(1- Nx) Now (0) = -12<0,f(I)=-11<0,/(2)=4>0 2-Nx,)x,, 2 =0,12, ~ @ So a real root of /(x)=0 lies between 1 and 2. Taking x, =15 for rapid convergence, we get net =n n= 0,12, which is the required iterative formula. 1 = Bd 412. 901-9)" 412 As the value of lies between 35 amd 3 aa 41-5)" e. between 0.04 and 0.08, s0 we take N=22%=00 Then, from (1),we have i 4 x, = (2-22 0,04)004 = 0048 a, = S208) +98 2 ay = (2-22 x 0.0448)0.0448 = 00454- 4(2-0138)° x) = 004545 = 18777 4 = 0.04545.NUMERICAL METHOD Theo u 5 ‘qPhus the required value of gy is 0.045, significant figures. Ex.14. Evaluate JT to three places of Raphson method. [W.B.U.T. CS-312, 2007, Mics, 9301, 49 Solution. Let x= J12 wx -12=0 Let f(x) = x*-12 2 f(a) =2x So the Newton-Raphson's formula f(x.) =x, - 2), n=0,1,2, Fut Fz) gives 12 2x, prantle| n=0,1,2,... 2, Now f(3)=-3, f(4)=4 So a root lies between 3 and 4 Choose x, = 3. 278 (3.5)° +12 2735 imilarly, x, = 3.46410 = 846410 = 3.46421 ame «. J12 = 3.464, correct upto 3 places of decimal. Exercise I. Snort Axswex Questions 1. Explain bisection method 2. Show that the convergence of bisection method is linear New, seornetric g.cive af 13.464 9.0.877 IL. LONG Answer Question: 1. Compute a real root of the following eyy aethod correct to five significant figures: 7 bition (@) x -32-5=0 (b) x -x-10=0 (© 2 -9x41=0 in [2,3] (8) cos.x = xe* (©) tanx+x=0 (9 38x—-cosx—1=0 in [0,1] 2. Using the method of bisection to compute a root of * 42-120 between 2 and 3 upto four significant digits 2, ‘ind the positive real root of x'-x'-1=0 using the bisection method of 4 iterations. [W.BU.T, CS-312, 2010)220 NUMERICAL METHOD Theoretical g ~ 4. Using the method of bisection to compute 4 31 p12 0 correct upto two significant digits, ot a | x WBULCS.319 5 5, Use Regula-Falsi method to evaluate the smallest ye "0, of each of the following equations Rong (a) 2 4+x°-1=0 (b) 8 -4x4+150 | (©) 2x°-3x-6=0 (©) 8-3x-5 WBU.T,CS.319 2m 6. Find out the root of the following equation usin, Falsi method: 3x~cosx~1=0 that lies between 0 and 1 correct (@) xe* = cosx i Regu, t decimal places) WB.UD, Cs.312, 200% 2008 1. Using Newton-Raphson method, find a real root of following equations correct to three decimal places @ xt-x-1=0 Gi) 3x +2x-9=0 | + Gil) 2x-3sinx-5=0 (ivyxt+e* =0 8. Construct a iterative formula to evaluate the following using Newton-Raphson method and hence evaluate | ONS GIR GY Gy YR | 9. Prove that Newton-Raphson’s iteration formula for Wis 1 N Fy = a “4. Hence find 3]. 10. Find the Newton-Raphson iteration formula to find the?" root of positive number N and hence find the cube root of 17. (W.B.U.T.,CS-312, 2009] 1 1 ¥ . = =,(2- Nx,). Hence find the value of Fg conrect upto two significant figures. 11. Prove that Newton-Raphson's iteration formula for x, 12. Find the root of ‘ e*—x=0, near x=0 correct to three significant figures, [W.B.U.7.,CS.312, 2002) Answers 92278 MISS 295 yg " 9 20289-00607 sin 3.2687 Seer 413 5.07548 0284 Gi) O72 ayy 5199 goat 70) 1221 $30 GH) 2885 Gy 569 5.09 135106 CD 1993 Gi) 2.7589 Gyr gogg 10,2513 1a Monies Cuotce Questions seendental equation may have oj atean ¥ have either no root or one or mor (a) True (b) False 2. Which of the following does not always guarantee the convergence (a) Bisection method (b) Regula-Falsi method (©) Newton-Raphson method (@ None of these (MAKA.U-T,M(CS)-401, 2016) 3. The rate of convergence of Newton-Raphson method is (0) linear (b) quadratic (©) cubic (d) none of these 4. The order of convergence of the Newton-Raphson’s method is @o i 2 @3 5. The Newton-Raphson’s method is fail when (@) f(a)=1 @) f'(x)=0 as © ren ur,M(CS)-401, 2016)ay [Oe MP” , ay {£09} Moog} wo {roo} IA 8, Ono of the roots of x" = 17x (a) Lai 9, One of the roots of x () Land 2 (c) Zand 3 11, Newton-Raphson’s method is also known as (a) normal method (b) tangent method (©) Parallel method (d) none of these 12, The rate of convergence of bisection method is: (b) quadratic (c) cubic (d) (W.B.U.1,C8-312, 2007, M(CS)-801, 20151 13, Binection method used for finding t no one of these 1 real root of @ (b) graphic (d) none of th (©) iterative mn sat. SOLUTIONS OF ALg, pric’ a wot rapilly bY N-Te method of the root is 15, Newton-Reaph Miata positive re x, i x, R =X,-— (6) Fmt wn 1 R (otal Fg 1 R (A) Fn = Bf Be 17, Regula-Falsi method is (a) * (a) Conditionally convergent () linearly convergent (©) divergent (d) none of these (W.B.U.T.,CS-312, 2009, 2013, M(CS)-401, 18 Regula Falsi method used for finding the real rts of Aumerical equation is () an 2014, 2016) (b) graphical method (d) none of these (WW.B.U.T,C8-912, 2004, 2006) nalytical method ethod (0) iterative m224 NUMERICAL METHOD Theoretical & Prag, 2 19, Newton-Raphson method even does not fail when f1 in the neighbourhood of the xeal root (a) True (b) False 20. If f(a)f(by) <0 and do 1 © |e @J=1 @ [6@)=2 [M.A.K.A.U.T.,.M(CS)-401, 2013) Answers la 2b 3.b 4c Sb 6c 7b 8 be 9D Ine ILb 12a 13.c 14b 15.c 16.c 17.2 18.6 19.b 206 2b 22.0 23.0 NUMERICAL Souunoy or Dirrerey ny, OnDwWany iptroanction: if . ial equation are ifferential © involved pifefang and Science. In this chant ™2%Y problems in Meal methods for solving ordinary air’ 248s Various js to study the solution of ferential equations ue ial value problem ows aim the initi dy Fe 7 LOY, I =H(x) » syhich_ (C8 9) 48 a continuous funtion of iawtin D of the xy-plane and (x,y) is a ghen pn me dameondition yp ~ f(%) is known as the initia an Sufieiont conditions for the existence and uniquenee of a Station of the equation (1) are the well known Linchin fonditions given by hite ( f+) is defined and continuous in D, the region containing (Xo, Yo) (ii) there exists a constant L such that [ro1)-132)|$ Abi yl ~ ® forall (x, 94) (92) €D- We now proceed to consider numerical techniques for solving (at a sequence of points x; = x9 + ih, called the mesh points, hbeing the step length. Let 3, be the approximation to the exact solution (.x;) of (1). A continuous approximation to y is then obtained by interpolating the data points ( . Euler’s method. aill@ shall now describe a method, [heh gives the solution in the formo! amine’ ‘step method, we determine tuck. a ‘step length) depending on e+) = y(x) + h9(x, 9h) +0(0""") veers (a) + hala, x50) +0(H"") Is known as Euler's method, fa set of tabulated values. ‘a fanction #58) 05.9 (a, 9) and its derivatives @) method, a positive integer, called the order of