JACOBI MATRICES FOR MEASURES MODIFIED BY A RATIONAL FACTOR. This paper decribes how, given the Jacobi matrix J for the measure d(t), it is possible to

Size: px
Start display at page:

Download "JACOBI MATRICES FOR MEASURES MODIFIED BY A RATIONAL FACTOR. This paper decribes how, given the Jacobi matrix J for the measure d(t), it is possible to"

Transcription

1 JACOBI MATRICES FOR MEASURES MODIFIED BY A RATIONAL FACTOR SYLVAN ELHAY AND JAROSLAV KAUTSKY y Abstrct. This pper decribes how, given the Jcobi mtrix J for the mesure d(t), it is possible to produce the Jcobi mtrix ^J for the mesure r(t)d(t) where r(t) is quotient of polynomils. The method uses new fctoring lgorithm to generte the Jcobi mtrices ssocited with the prtil frction decomposition of r(t) nd then pplies previouly developed summing technique to merge these Jcobi mtrices. The fctoring method performs best just where Gutschi's miniml solution method for this problem is wekest nd vice vers. This suggests hybrid strtegy which is believed to be the most powerful yet for solving this problem. The method is demonstrted on simple exmple nd some numericl tests illustrte its performnce chrcteristics. Key words. orthogonl polynomils, Jcobi mtrices, rtionl fctors modifying mesures, Guss qudrtures 1. Introduction. Let d(t) 0, mesure with nite, innite or semi-innite support [; b], be such tht the power moments (1) j = t j d(t) exist nd re nite for ll j = 0; 1; 2;. Let J be the (symmetric, tri-digonl) Jcobi mtrix, the elements of which re the coecients in the three-term recurrence reltion for the polynomils p j (t) j = 0; 1; 2; which re orthonorml with respect to d(t) on [; b]. Jcobi mtrices nd their ssocited orthogonl polynomils rise in connection with pproximte numericl integrtion, lest squres pproximtions for continuous nd discrete mesures, series expnsions nd continued frctions. The Jcobi mtrix for some d(t) cn, in principle, be computed from the power moments (1), but the build-up of roundo errors in this process is so severe tht in prctice this method is lmost useless. However, for certin of the clssicl mesures the Jcobi mtrices re explicitly known (see [2],[10] for some exmples) nd much eort hs gone towrds the investigtion nd development of methods which cn be viewed s trnsformtions from one or more known Jcobi mtrices, J 1 ; J 2 ;, corresponding to mesures d 1 (t); d 2 (t);, into the required Jcobi mtrix ^J for the mesure d^(t) which is relted to the d i (t). This is, of course, essentilly the sme s trnsforming one set or sets of orthogonl polynomils into set of others, but is done in numericl liner lgebr context. Thus, trnsformtion methods bsed on the modied moments (2) ^ j = p j (t)d^(t); j 0; (in which the p j re orthogonl with respect to the mesure d(t)), were proposed by Sck nd Donovn in [9]. The Lower Tringulr Lnczos (LTL) method of [8] (which is equivlent to the Modied Chebyshev method of [5]), is bsed on modied moments nd is of prticulr interest in this pper. Computer Science Deprtment, University of Adelide, South Austrli, y School of Informtion Science nd Technology, Flinders University, Bedford Prk, South Austrli, Pge 1, August 18, 1993

2 In [8], Kutsky nd Golub show, for r(t) 0 8t 2 [; b] polynomil, how to nd ^J for d^(t) = r(t)d(t) from J for d(t) (Polynomil Shift Implicit QR or PSIQR method). The present uthors used this pproch [7] to extend this work to the cse where the polynomil r cn chnge sign in [; b] nd use the methods so developed [4] to compute sequences of Kronrod-Ptterson type imbedded qudrture sequences for two clssicl mesures, e t on [0; 1] nd e t2 on [ 1; 1], for which Jcobi mtrices re explicitly known. More recently Elhy, Golub nd Kutsky show tht it is possible (see [3]), given the Jcobi mtrices J 1 nd J 2 for the mesures d 1 (t) nd d 2 (t), to produce the Jcobi mtrix J 3 corresponding to the mesure (3) d 3 (t) = 1 d 1 (t) + 2 d 2 (t); 1;2 constnts, if such mtrix J 3 exists (SUM method). Gutschi, in [5] nd [6], shows how to compute the modied moments of the mesures (4) nd d(t)=(t v); v =2 [; b]; (5) d(t)=((t x) 2 + y 2 ); x 2 R; y > 0; by mens of so-clled miniml solution to the three-term recurrence for the polynomils orthogonl with respect to d(t) (MS method). Once these modied moments re vilble it is possible to compute the corresponding Jcobi mtrix by mens of the LTL method. Note tht the denomintor in (5) is just (t v)(t v) with complex v = x+iy. Thus modictions such s the ones in (4) nd (5) llow for computtion of J corresponding to d(t) divided by either rel liner fctor or pir of complex conjugte fctors. Erlier work by Uvrov [11] uses the second solution nd Christoel-Drboux reltions to estblish determinntl expressions which relte the polynomils orthogonl with respect to some d(t) to those orthogonl with respect to r(t)d(t), r(t) rtionl frction with known fctors. However, there seems to be no esy wy to use them to determine the required Jcobi mtrix. The PSIQR nd SUM methods re bsed on orthogonl rottions nd re numericlly very stble. While the MS cn produce very ccurte Jcobi mtrices for mesures of the type (4) nd (5) it suers from the problem tht, for mny importnt cses, infesibly lrge initil mtrices my be required in order to produce quite moderte dimension result. In these cses the problem is compounded if we need to pply the MS method to (lredy divided) mesures such s (4) nd (5) repetedly, for exmple, to produce Jcobi mtrix tht corresponds to d(t) divided by product of fctors. Thus, there exist very stisfctory methods for Jcobi mtrices tht correspond to sums nd dierences of mesures, or products of mesures with polynomils, but there remins need for stisfctory methods to compute the Jcobi mtrix for mesure modied by quotient of polynomils. More precisely, we re interested in the following problem: Problem 1.1. Let the rtionl function r(t) = g n (t)=h m (t), g n nd h m polynomils of degree n nd m respectively, be such tht h m hs no zeros in [; b] nd r(t) 0, 8t 2 [; b]. Given nite J for d(t) nd ^J for d^(t) = r(t)d(t). Pge 2, August 18, 1993

3 Clerly, since d^(t) is non-negtive on the intervl [; b], there exist polynomils ^p j (t), j = 0; 1; 2;, ech of exct degree j, which re orthogonl on [; b] with respect to the inner product (f; g) = f(t)g(t)r(t)d(t): The importnce of stisfctory solution to Problem 1.1 is illustrted by the following ppliction. Suppose we know J for d(t) nd we require the Jcobi mtrix ^J for d^(t) = q(t)d(t), q some smooth function. If the intervl of orthogonlity is nite we cn pproximte q by polynomil g n nd use the PSIQR method to trnsform J into ^J for d^(t) = g n d(t). But on n innite intervl no polynomil cn usefully pproximte smooth function which does not grow beyond ll bounds t innity. Similrly, function with singulrity close to the support of d(t) my not be well pproximted by ny polynomil. This is just where rtionl pproximtions cn be used with good eect. Thus Jcobi mtrices hve importnt ppliction in the computtion of Gussin qudrture formule for mesures of the type d^(t) = r(t)d(t) where r is rtionl pproximtion to q. In the next section we derive new lgorithms which trnsform the known Jcobi mtrix J for the mesure d(t) into the Jcobi mtrix ^J corresponding to d(t) divided by liner or complex conjugte pir qudrtic fctor. We cll these, collectively, the Inverse Cholesky (IC) method. In x3 we indicte the IC lgorithms themselves nd briey review the MS method. The discussion in x4 leds to hybrid strtegy for hndling rtionl modiction which uses both the IC nd MS methods. In x5 we show simple exmple nd in x6 we summrize the results of some numericl tests. In x7 we mke some concluding remrks. Tbles of numericl results nd certin moments, which my be needed when using the IC method with the clssicl mesures, re displyed in x8. 2. The new methods. In x1 we mentioned tht the Jcobi mtrix for mesure modied by multipliction with polynomil is stbly nd eciently done with the PSIQR method. So our concern in the solution to Problem 1.1 is in the determintion of Jcobi mtrix tht corresponds to mesure d(t) modied by division with polynomil h m (t). We ssume tht the m (rel or complex conjugte pir) zeros of h m re known nd simple. We now estblish some nottion nd derive the reltions on which the IC method is bsed. Suppose tht the polynomils p(t) = (p 0 (t); p 1 (t); ; p n 1 (t)) T nd p n (t) re orthonorml on the intervl [; b] with respect to the non-negtive mesure d(t). Then there exists (symmetric, tridigonl) mtrix J = C C... A ; n the elements of which re the coecients in the three-term recurrence which the polynomils stisfy. In the following identity, s elsewhere in this pper, (6) tp(t) = Jp(t) + n p n (t)e n Pge 3, August 18, 1993

4 the vector e n is the lst column of n identity mtrix of pproprite dimension. Suppose now tht ^p(t) = (^p 0 (t); ^p 1 (t); ; ^p n 1 (t)) T is second set of polynomils orthonorml with respect to some mesure d^(t). The orthogonlity of the polynomils implies (multiply (6) on the right by p(t) T nd integrte) (7) p(t)p(t) T d(t) = I; tp(t)p(t) T d(t) = J nd there re corresponding reltions for the second set of orthonorml polynomils. Furthermore, there exist nonsingulr lower tringulr mtrix L, vector c nd non-zero constnt n, ll such tht (8) p(t) = L^p(t); p n (t) = c T ^p(t) + n ^p n (t): Thus, if ^J is the Jcobi mtrix in the identity corresponding to (6) for the polynomils ^p(t), ^p n (t) it is esy to show [4] tht (9) ^J = L 1 J L + e n b T ; where L 1 e n = n e n for some non-zero constnt n nd b = n n c. If we need the mtrix ^J of dimension n we cn get it by discrding the lst row nd column of the n + 1 dimension product L 1 JL, thus obviting the need to compute b. The methods we present in this pper determine the mtrix L directly from J nd some dditionl informtion. Our strting point is the theorem: Theorem 2.1. Let r(t) be ny nlytic function of t. Then (J ti)p(t)r(t) = (J ti)r(j)p(t) + n p n (t)(r(j) r(t)i)e n : If t is not n eigenvlue of J then (10) p(t)r(t) = r(j)p(t) + n p n (t)(j ti) 1 (r(j) r(t)i)e n : If v is such tht p n (v) = 0 then r(v) is n eigenvlue of r(j) nd p(v) is the corresponding eigenvector. Proof. From (6) we cn write (J ti)p(t) = n p n (t)e n : If r(t) is n nlytic function of t then J nd r(j) commute so we hve (J ti)(r(j) r(t)i)p(t) = n p n (t)(r(j) r(t)i)e n from which (10) immeditely follows. Multiplying (10) on the right by 1 p(t) nd integrting gives r(t) p(t)p(t) T d(t) = r(j) n p(t)p(t) T d(t) r(t) + p n (t)(j ti) 1 (r(j) r(t)i)e n p(t) T d(t) r(t) : Now if the polynomils p(t) re orthonorml with respect to the mesure d(t) nd the polynomils ^p(t) re orthonorml with respect to the divided mesure d^(t) = d(t) r(t) Pge 4, August 18, 1993

5 then we hve, fter substituting with (8), (11) I = r(j)ll T + n (c T ^p(t) + n^p n (t))(j ti) 1 (r(j) r(t)i)e n^p(t) T d^(t)l T : The two cses of present interest re r(t) = t v nd r(t) = (t x) 2 + y 2. For the cse r(t) = t v reltion (11) reduces to (12) I = (J vi)ll T + n (c T ^p(t) + n ^p n (t))e n^p(t) T d^(t)l T ; = (J vi)ll T + n e n c T L T by orthogonlity. Note tht the mtrix L T is Cholesky-like fctor of (J vi) plus rnk-one correction to its lst row. For the second cse of interest, r(t) = (t x) 2 + y 2, we hve I = ((J xi) 2 + y 2 I)LL T + (13) n (c T ^p(t) + n ^p n (t))((j xi) + (t x)i)e n^p(t)d^(t)l T : Now the second term on the right of (13) reduces, by orthogonlity, to ( ) n (J 2xI)e n c T + e n c T t^p(t)^p(t) T d^(t) L T T + ^ n e n e n (14) = n n(j 2xI)e n c T + e n c T ^J o L T + ^ n e n e n T for some constnt ^ n. Bering in mind tht the mtrix ^J in (14) is tridigonl, it is esy to see tht (14) is mtrix, the rst n 2 rows of which vnish, so the L T mtrix for this cse is Cholesky-like fctor of (J xi) 2 + y 2 I plus rnk-two correction to its lst two rows. The methods we describe in this pper directly generte the mtrix L, in ech of the bove cses, from the top down. If the process is terminted before reching the row or rows which re ected by the low rnk corrections, then the prt of L produced to tht point will be correct nd the low rnk corrections need not be computed. 3. The lgorithms. In this section we derive the lgorithms for the Cholesky-like fctors of the inverse of symetric tridigonl nd pentdigonl mtrices. In this section we write L i:j;k:m to denote rows i through j of columns k through m of the mtrix L. (15) 3.1. The liner-fctor Inverse Cholesky lgorithm. Given symmetric tridigonl mtrix J, we require the fctor L which stises where the lower tringle L hs elements L = 0 I = JLL T + e n d T ; `n1 `n2 `n3 : : : `nn Pge 5, August 18, C ` : : : 0 `21 `22 0 : : : 0 `31 `32 `33 : : : A :

6 Noting tht e T i J = ( 0 : : : 0 i 1 i i 0 : : : 0 ) 1 < i < n we hve, considering only 1 j i n, ( ij ni d j ) = i 1 n X = i 1 `i 1;k`jk + i n X min i 1;j X X + i j 1 `i 1;k`jk + i `ik`jk + i n X `i+1;k`jk jx `ik`jk `i+1;k`jk + i`i+1;j`jj : Algorithm 3.1, which is bsed on this reltion, determines the L nd d of (15). Its input is symmetric tridigonl mtrix J nd `11 6= 0. Algorithm = 0 `21 = (1=`11 1`11 )= 1 for i = 2 : n 1 `i+1;1 = ( i 1`i 1;1 + i`i1 )= i end d 1 = ( n 1`n 1;1 + n`n1 ) for j = 2 : n s = j 2 L j 2;1:j 2 L T j;1:j 2+ j 1 L j 1;1:j 1 L T j;1:j 1+ j 1 L j;1:j 1 L T j;1:j 1 `jj = p s= j 1 for i = j : n 1 s = i 1 L i 1;minfi 1;jg L T j;minfi 1;jg+ i L i;1:j L T j;1:j+ i L i+1;1:j 1 L T j;1:j 1 `i+1;j = ( ij s)= i =`jj end d j = ( n 1 L n 1;1:minfn 1;jg L T j;1:minfn 1;jg + n L n;1:j L T j;1:j) end 3.2. The qudrtic-fctor Inverse Cholesky lgorithm. Here we require the lower tringle L which stises the reltion (16) I = J 2 LL T + e n d T + e n 1 f T : where, J being tridigonl, J 2 is pentdigonl nd denoted (for dierent 's nd 's) by J 2 = C C... A : n We ssume tht the principl 2 2 block of L is prescribed. We hve e T i J = ( 0 : : : 0 i 2 i 1 i i i 0 : : : 0 ) ; 2 < i < n 2; Pge 6, August 18, 1993

7 nd denoting L s in the previous section, we hve, for 1 j i n, ( ij n;i 1 f j ni d j ) = i 2 minfi 2;jg X `i 2;k`jk + i 1 jx + i `ik`jk + i X + i j 1 jx minfi 1;jg X `i+1;k`jk `i+2;k`jk + i`i+2;j`jj : `i 1;k`jk Algorithm 3.2, which is bsed on this reltion, determines the L of (16). Its input is the symmetric pentdigonl mtrix J 2 nd the 2 2 principl submtrix of L. The lgorithm then nds the rest of L. The steps which compute d nd f re omitted for brevity. Algorithm 3.2. `3;1 = (1=`1;1 1`1;1 1`2;1 )= 1 `4;1 = ( 1`1;1 + 2`2;1 + 2`3;1 )= 2 for i = 4 : n 1 `i+1;1 = ( i 3`i 3;1 + i 2`i 2;1 + i 1`i 1;1 + i 1`i;1 )= i 1 end `3;2 = ( 1`2;1`1;1 + 1`22;1 + 1`2;1`3;1 + 1`22;2)=( 1`2;2 ) s = 1`2;1`1;1 + 2`22;1 + 2`2;1`3;1 + 2`2;1`4;1 + 2`22;2 + 2`2;2`3;2 `4;2 = (1 s)=( 2`2;2 ) for i = 3 : n 2 s = i 2 L i 2;1:minfi 2;2g L T 2;1:minfi 2;2g + i 1L i 1;1:minfi 1;2g L T 2;1:minfi 1;2g + i L i;1:2 L T 2;1:2 + i L i+1;1:2 L T 2;1:2 + i`i+2;1`t2;1 `i+2;2 = s=( i`2;2 ) end for j = 3 : n s = 0 if j > 3; s = s + j 3 L j 3;1:j 3 L T j;1:j 3 ; end if j > 4; s = s + j 4 L j 4;1:j 4 L T j;1:j 4 ; end s = s + j 2 L j 2;1:j L T j;1:j + j 2L j 1;1:j L T j;1:j + j 2L j;1:j 1 L T j;1:j 1 `j;j = p s= j 2 for i = j 1 : n 2 s = i 2 L i 2;1:minfi 2;jg L T j;1:minfi 2;jg+ i 1 L i 1;1:minfi 1;jg L T j;1:minfi 1;jg+ i L i;1:j L T j;1:j + i L i+1;1:j L T j;1:j + i L i+2;1:j 1 L T j;1:j 1 `i+2;j = ( ij s)=( i`j;j ) end end Strting vlues. The liner fctor IC method will be used for rel v. To strt it we need the vlue of (see (8)) where `11 = (~ 0 = 0 ; ) 1 2 d(t) ~ 0 = t v ; Pge 7, August 18, 1993

8 is vilble, from the tbles provided, for the clssicl mesures. The qudrtic fctor IC method will be used for complex conjugte pirs v, v where v = x + iy, x; y rel, y > 0. To strt it we need the principl 2 2 submtrix of L, `11 0 : `21 `22 The mesure d(t) hs here been modied by the fctor (t v)(t v) = (t x) 2 + y 2. Now denote t ~ j = j (t x) 2 d(t); j = 0; 1; 2: + y2 Using (8) nd orthognlity it quickly follows tht `2 11 = ~ 0 = 0 `21 = `11 (~ 1 =~ 0 1 )= 1 `2 22 = (~ 2 ~ 2 1 =~ 0)=( 0 2 1): For the clssicl mesures, the ~ j cn be found, gin in terms of the tbled quntities, by ~ 0 = ~ 1 = iy Z 1 b t v d(t) Z 1 b t v d(t) + ~ 2 = 0 + 2x~ 1 (x 2 + y 2 )~ 0 :! 1 t v d(t) 1 t v d(t)! x~ Modied moments from the miniml solution. For completeness, we stte here Gutschi's miniml solution (MS) method re-written to operte for normlized rther thn monic polynomils. Given J of dimension m, the zero-th moment, 0 of d(t), knot v outside the support intervl of d(t) nd n integer n = n(m; v) < m, the lgorithm produces the vector ^ = ( 0 ; 1 ; 2 ; : : : ; n 1 ) T. For rel v, the modied moments of the mesure (4) re the ^ i = i nd for complex v the modied moments of the mesure (5) re given by ^ i = Img( i) Img(v) : Algorithm 3.3. p m = 0 for i = m : 1 : 2 p i 1 = i 1 =(v i i p i ) end 0 = p 0 =( 1 p 1 (v 1 )) for i = 1 : n 1 i = i 1 p i end Pge 8, August 18, 1993

9 For the mesures in Tble 5 Gutschi proposes lower bound on n(m; v) which is lrge enough to ensure tht the ^ i re computed to suciently high ccurcy. As the point v pproches support of d(t) the vlue of n(m; v) increses nd lrger dimension J is needed for the sme ccurcy. 4. Use of the IC nd MS methods. Our discussion of just how to use the IC nd MS methods divides nturlly into two prts. The rst prt ddresses the question of how to use the methods repetedly to nd the J for mesure divided by generl polynomil of degree m > 2. The second prt ddresses the question of how to decide on which of the IC nd MS methods should be used for ech of the liner nd qudrtic fctors in the rtionl function denomintor. This issue rises becuse, s we show lter in the numericl tests, the IC nd MS methods behve very dierently from one nother ccording to whether the division fctors correspond to poles close to, or fr from, the support of the mesure d(t) Repeted divisions. In this section we compre () the strtegy of strting with known J, modifying it by division with fctor, modifying the result by division with the next fctor, nd so on, (the product strtegy), with (b) the strtegy of modifying the known J for division by ech of the fctors seprtely nd then using the summing method to produce the nl result (the summing strtegy). To strt, we compute the prtil frction decomposition of h m (t) = my (t v k ): In the cse where the zeros of h m re conjugte pirs, the prtil frction decomposition contins terms of the form (17) Ax + B (t x) 2 + y 2 ; y > 0: To compute the Jcobi mtrix for such term we use one of the lgorithms below for the denomintor rst nd then pply the PSIQR method to the result for the numertor The product strtegy. Denote by J s the Jcobi mtrix for the mesure (18) d(t) Q s (t v k) : In the product strtegy we strt from the given J 0 = J for d(t) nd sequentilly clculte J 1 ; J 2 ; : : :, nishing up with J m = ^J for d(t)=hm (t). Ech step in this sequence, in which we produce J s from J s 1, cn be performed, in principle, by either of the following lgorithms (for complex pirs of zeros the IC method cn be used to trnsform from J s 1 to J s+1 directly using only rel rithmetic, s indicted in x3, but the MS method requires complex oting point rithmetic). One stge using the MS method requires n lgorithm such s: Algorithm 4.1. Input: J s 1 of suciently lrge dimension j nd strting vlues s in x Pge 9, August 18, 1993

10 () Use the MS method to produce 2(i + 2) modied moments. (b) Use the LTL method to directly get J s of dimension i. By comprison, one stge using the IC method would require Algorithm 4.2. Input: J s 1 of dimension i+1 nd strting vlues s in x () Use the IC method to produce L s of dimension i + 1. (b) Compute Ls 1 J s 1L s = J s e n b T s of dimension i + 1 (see (9)). (c) Discrd the lst row nd column of the product to get J s of dimension i. However, there re diculties with Algorithm 4.1. The estimtes of the size of j for given v k when d(t) is one of the clssicl (Jcobi, Lguerre or Hermite) mesures cn be quite lrge, if the point v k is close to the support of the intervl: for exmple to modify dimension i = 10 Lguerre Jcobi mtrix by the fctor (t 0:01) requires n initil Jcobi mtrix of dimension nerly j = 2000 when using stndrd double precision IEEE rithmetic. Thus, even for quite modest nl dimension ^J, very lrge initil mtrix my be needed. More problemtic is the fct tht the only estimtes for the size of the initil mtrix t ech stge re for the clssicl mesures nd there re not estimtes for the divided clssicl mesures. It is possible tht one could check the ccurcy of the process t ech stge by recomputing with vried estimtes of the strting size mtrix but clerly this process is unstisfctory. Indeed, it my even be necessry to bndon the clcultion nd strt gin from scrtch t some stge if the remining Jcobi mtrix is too smll to go to the next step. Using the Algorithm 4.2 requires n initil mtrix of only bout n + m but there re diculties s before: the strting vlues, which re computed from the zero-th moments of mesures such s (18) re not generlly vilble The summing strtegy. An lterntive strtegy uses the results of [3] mentioned in x1. If we denote the method for computing the Jcobi mtrix which corresponds to sum or dierence of mesures by SUM then the lgorithm we propose is: Algorithm 4.3. Given J of dimension n + 1 for d(t) () For ech k = 1 : m either (i) Use the MS method to produce 2(n + 2) modied moments for J ~ vk corresponding to d~(t) = d(t)=(t v k ) (or d~(t) = d(t)=(t v k )(t v k )). (ii) use the LTL method with these modied moments to get J ~ vk of dimension n for d~(t) or (i) Use the IC to produce L k for d~(t) = d(t)=(t v k ) (or d~(t) = d(t)=(t v k )(t v k )). (ii) Compute L 1 k JL k = J ~ vk e n b T v k of dimension n + 1 (or n + 2) (see (9)). (iii) Discrd the lst row (two rows) nd column (two columns) of the product to get J ~ vk of dimension n for d~(t). (b) Use method SUM to nd ^J from the seprte Jcobi mtrices. This scheme overcomes the disdvntges, mentioned bove, of the product strtegy since the strting vlues re known (for both the IC nd MS methods) for the clssicl mesures nd, in the cse of the MS method, ll the divisions use the sme initil mtrix nd its dimension cn be determined once t the outset from the estimtes given in [5]. If the numertor of ny term such s (17) chnges sign in the support of d(t), there my not exist Jcobi mtrix for it. However, by dding nd subtrcting Pge 10, August 18, 1993

11 constnt to the numertor we cn rewrite (17) s two seprte terms ech of which does not chnge sign on the support. We cn then sum ll the terms which re positive on the support nd then subtrct those which re negtive. Since the nl Jcobi mtrix corresponds to mesure which is positive on the support (by ssumption), the result must exist nd be computble this wy Distribution of the poles. Consider the cse of modifying J for d(t) to ^J for d^(t) = d(t)=(t v), v rel or complex, outside the support, [; b], of d(t). The closer is v to [; b], the lrger j, the dimension of J, needs to be for the MS method to mintin ccurcy. The estimtes given in [5] for the size j when d(t) is one of the Jcobi, Lguerre or Hermite mesures, re reported there to be relistic in the sense tht, using J of dimension 5 fewer thn given by the estimte rrely gve sucient ccurcy nd the given estimte ws only occisionlly required to be incresed by bout 5 or 10. Thus close to the support, the MS method cn be quite unstisfctory. By contrst, the IC method, s the numericl tests reported in x6 show, returns poorer results s the point v moves wy from the support but is extremely ccurte close to the support. Our numericl evidence lso suggests tht this eect is more pronounced for the mesures with innite or semi-ininte support. The obvious solution then, is to use Algorithm 4.3, pplying the MS method on points fr from [; b] nd the IC methods on those close to [; b]. Both methods cn be pplied to points which re neither close nor fr nd the two results used for conrmtion. Thus, Algorithm 4.3 provides stble nd ecient method for solution to Problem A simple exmple. The function ln(1 + 4t) hs power series ln(1 + 4t) = 4 t 8 t 2 + nd Pde pproximtion (19) 64 t t t t6 3 r(t) = g 3 (t)=h 3 (t) = 60 t t t t t t t7 7 + O(t 8 ); which, on the subintervl [0; 20], hs error bounded, in bsolute vlue, by bout 0:228. The denomintor h 3 (t) hs zeros f 1=2; 1 4 (5 p 15)g f 0:5; 0:28175; 2:21825g nd the numertor g 3 (t) hs zeros f0; :3; 1g so both g 3 (t) nd h 3 (t) do not chnge sign on the support [0; 1] of the Lguerre mesure. The prtil frction decomposition of 1=h 3 (t) is A 1 + A 2 + A 3 t v 1 t v 2 t v 3 = 1=18 t + 1= =(15 3 p 15) 6 (t + (5 p 15)=4) =( p 15) (t + (5 + p 15)=4) : We demonstrte the technique by nding the dimension 7 Jcobi mtrix which corresponds to the mesure d(t) = e t r(t) on the intervl [0; 1] from the (known, see [2]) Jcobi mtrix for the mesure d(t) = e t on the sme intervl. Since Pge 11, August 18, 1993

12 the processes we employ require us to discrd one or two rows nd columns of the computed mtrices, we ssume tht the strting mtrix is suciently lrge to give result of the required dimensions. All the clcultions were done in IEEE stndrd double precision rithmetic (mchine epsilon 2: ). Let us denote the following power moments by Z 1 t ~ jvi = j e (20) t dt; i = 1; 2; 3; j = 0; 1; 2; : t v i 0 To strt with, we use Tble 6 to compute ~ 0vi for ech of the zeros v i. These numbers (computed to 50D with Mple V nd correctly rounded) re given in Tble 1. Of v 1 9: v 2 1: : v 3 Tble 1 Vlues of ~ for the three zeros of h 0v i 3(t). course, 0 = 1. We next use Algorithm 4.3 (choosing the IC method option) to compute the three Jcobi mtrices J ~ vi, i = 1; 2; 3, ech of which corresponds to the mesure e t =(t v i ). We now pply the method SUM, referred to in x1, to J ~ v2 nd J ~ v3 to compute the Jcobi mtrix, J temp1, corresponding to the sum of the two positive terms e t A 2 =(t v 2 ) nd e t A 3 =(t v 3 ). The constnts i, i = 1; 2 of (3) re here chosen s A i ~ 0vi =(~ 0v2 + ~ 0v3 ). Next, we use the dierencing version of SUM to compute the Jcobi mtrix, J temp2 corresponding to the mesure e t A2 t v 2 + A 3 t v 3 + e t A 1 t v 1 : (recll tht A 1 < 0) from J temp1 nd J v1. Finlly, we pply the PSIQR method (see x1) to J temp2 with the polynomil g 3 (t) to get ^J corresponding to r(t)d(t). After the PSIQR process we discrd the lst 2 rows nd columns of the mtrix ^J. The resulting mtrix hs digonl j nd subdigonl j components shown in Tble 2 nd the 7-point Guss qudrture formul (21) Q(f) = nx j=1 w j f(t j ) found from it is in Tble 3. Of course, this Jcobi mtrix determines ll the Guss qudrtures (for the modied mesure r(t)e t ) with seven or fewer points. It is worth noting tht we could hve run the processes in dierent order: strt with the Jcobi mtrix for e t, pply the PSIQR method with polynomil g 3 (t) nd pply the IC method to this mtrix once for ech of the poles of r(t). The three resulting mtrices could then be merged using the SUM method. Although it is not relible test of the ccurcy of qudrture fromul, we pplied Q(f) to the powers of t to conrm tht it is indeed the required formul. We expect Q(f) to compute the integrls (22) I pprx (t j ) = Z 1 t j e t r(t)dt 0 Pge 12, August 18, 1993

13 j j j 1 1: : : : : : : : : : : : : Tble 2 The Jcobi mtrix for r(t)e t. j t j w j 1 3: : : : : : : : : : : : : : Tble 3 The 7-point Guss qudrture formul for r(t)e t. to within modest multiple of mchine ccurcy for the rnge j = 0; 1; 2; : : :; 13 nd we expect I pprx (t j ) Q(t j ) to be quite dierent from (mchine) zero for j = 14. Further, we expect tht Q(f) will pproximte (23) I exct (t j ) = Z 1 0 t j e t ln (1 + 4t)dt to roughly the ccurcy tht I pprx (t j ) pproximtes I exct (t j ). In fct, I pprx (t j ) pproximtes I exct (t j ) with reltive error tht vries lmost linerly between nd s j rnges from 0 to 14. Once ~ 0vi is known, the numbers ~ jvi, j > 0 cn esily be computed from the recursion (24) ~ jv = v~ j 1;v + j 1 ; j > 0: From these nd the prtil frction decomposition, we cn nd the I pprx (t j ) of (22). Furthermore, the numbers j = I exct (t j ) of (23) cn be found from the recursion j = j j 1 + ~ j; 1 ; 0 = ~ 4 0; 1 4 once (24) hs been used with v = 1=4 to compute the ~ j; 1. The bsolute vlues of the reltive errors for Q(t j ) pproximting the numbers in (22) nd (23) re 4 summrized in Tble 4 nd follow our expecttions. 6. Numericl Tests. In this section we report the results of some tests which compre the performnce of the IC method nd the MS method nd which identify two signicnt dierences in the wy these methods behve. Our tests were pplied to ll the clssicl mesures shown in Tble 5. In ll cses the clcultions were done Pge 13, August 18, 1993

14 Error of Q(t j ) for j I exct (t j ) I pprx (t j ) 0 1: : : : : : : : : : : : : : : : : : : : : : : : : : : : : : Tble 4 Reltive errors of Q(t j ) in IEEE stndrd double precision rithmetic (mchine epsilon 2: ). The tbles in this section show the ccurcy of computed Jcobi mtrices when compred with the explicitly known results. For ech rel knot v, or complex knot x + iy, the entry heded N in the tble shows the number N = n(m; v) of recurrence coecients which the MS method requires. Where this number ws considered too lrge to be fesible we used 2000 coecients nd disply the required N in prentheses. The remining columns show the ccurcy of the digonl j nd subdigonl j elements produced by the IC nd MS methods. Thus for the columns heded we show ( min j j exct log 10 pprox j exct j nd, except where they re zero by virtue of the symmetry of the mesure d(t), similr quntities for the digonls j. Recll tht the IC method produces mtrix L from which we compute the required Jcobi mtrix using (9). We computed the 2-norm condition numbers for the L mtrices tht re generted by the IC process for these tests nd found tht they remin modest for ll the cses shown. For n = 10 ll but 5 re smller thn 80, nd the blnce re smller thn 700. For n = 50 they re ll smller thn The tests themselves exploit the fct tht the PSIQR nd either the IC or MS re inverses of ech other if the polynomil r(t) is rel liner fctor (r(t) = t v) or the product of complex conjugte pir of fctors (r(t) = ((t x) 2 + y 2 ), y > 0). Since the PSIQR method produces mtrices which re ccurte to bout mchine epsilon, the dierence between the strting mtrix nd the result, fter pplying both the PSIQR nd either IC or MS, is good mesure of the error. This error is just wht we disply in the tbles of this section. The tbles illustrte tht the MS method performnce improves s the point v moves wy from the support of d(t) nd lso shows tht the MS method is better overll on mesures which hve nite support. By contrst, the IC method Pge 14, August 18, 1993 ) :

15 performnce improves s the point v pproches the support intervl nd is better overll for the mesures with innite support. These phenomen re clerly shown s the point v pproches the support intervl from the left, in the cse of liner fctors (Tbles 7 nd 8) nd s it pproches the origin long the line y = x in the cse of complex pir fctors or for the Hermite mesure (Tbles 9, 10 nd 11). Tble 12 shows the errors for the methods when the point x + iy moves prllel but close to the x-xis from x = 5 to x = 5 nd illustrtes further the behviour of the two methods close to nd fr from the support. Most of the tbles in this section show the results of the tests pplied to produce mtrices of dimension n = 10. However, these tbles re representtive of lrger cses s the exmples in Tbles 13 nd 14, with n = 50, show. The loss of ccurcy seems to be quite grdul s n increses. We used the non-symmetric Jcobi mesure (with mesure prmeters = 1=2, = 1=7) s n exmple of nite support mesure to illustrte the performnce of the methods on both the j nd the j components of the mtrix. The Legendre, Gegenbuer nd Chebyshev mesures give similr results. The Hermite nd Lguerre mesures used here hve the mesure prmeter = 1=3. The clcultion seems, however, to be reltively insensitive to this prmeter nd other vlues of tht we tested gve very similr results. These dierences in performnce suggest hybrid method for the solution of Problem 1.1. The decision bout which of the IC or MS methods to choose in Algorithm 4.3 should be driven by the phenomen the results here indicte: the closeness of the point in question to the support intervl nd whether or not the suport intervl is innite or semi-innite. Where either method is suitble they could both be used nd the results compred for veriction. 7. Conclusions. We hve derived new method, clled the IC method, for the computtion of the Jcobi mtrix corresponding to mesure modied by rtionl function. Our method is bsed on new fctoring lgorithm which cn be used to produce the Jcobi mtrix corresponding to known mesure divided by liner or qudrtic fctor. We use the fctoring lgorithm once for ech pole or conjugte pir of poles nd then pply the uthors' SUM method to nd the Jcobi mtrix corresponding to the given mesure divided by the denomintor of the rtionl function. We then pply the PSIQR method to this for the numertor. The IC method requires certin strting vlues which re expressed, for the clssicl mesures, in terms of well known specil functions. We computed these to very high ccurcy without diculty by use of the symbolic mnipultors Mple nd Mthemtic. Subroutine pproximtions for these specil functions re lso redily vilble from softwre librries such s Netlib. We hve shown tht the product strtegy of x4 hs serious problems nd tht our new summing strtegy overcomes these problems. We hve lso shown tht the IC method performs best just where the previously known MS is wekest nd vice vers. This hs led to hybrid strtegy which we believe is the most powerful yet for this problem. The new method ws demonstrted on simple exmple nd some tbles, which re representtive of the numericl tests we rn, show the pttern of performnce. Problem 1.1 remins dicult nd stble method which does not require the poles of the rtionl function to be known is still desirble. 8. Tbles. Pge 15, August 18, 1993

16 8.1. Zero-th moments of clssicl mesures nd divided clssicl mesures. The following re sometimes referred to s the clssicl mesures: Nme d(t) Intervl Constrints Legendre 1 [ 1; 1] Chebyshev (1 t 2 ) 1=2 [ 1; 1] Gegenbuer (1 t 2 ) [ 1; 1] > 1 Jcobi (1 t) (1 + t) [ 1; 1] ; > 1 Lguerre t e t [0; 1] > 1 Hermite jtj e t2 [ 1; 1] > 1 Tble 5 Clssicl mesures. The tble below shows the zero-th moments, 0, of these mesures nd the zero-th moments of these when divided by liner fctor, ~ 0 = 1 t v d(t); v = x + iy 62 [; b]: In the cse of the Hermite mesure we require y > 0, but v my be rel for the other mesures. Nme 0 ~ 0 Legendre 2 ln((v 1)=(v + 1)) Chebyshev = p v 2 1 Gegenbuer Jcobi (+1) 2 (2+2) (+1) (+1) (++2) (1 v) B(1 + ; 1 + ) 2F 1 (1; + 1; 2 + 2; 2 1 v ) (1 v) B(1 + ; 1 + ) 2F 1 (1; + 1; + + 2; 2 1 v ) Lguerre ( = 0) 1 e v E 1 ( v) Hermite ( = 0) p Tble 6 Moments of clssicl mesures. ie v 2 erfc( iz) Here the Gmm, Bet, nd Hypergeometric functions re dened by (z) = Z 1 t z 1 e t dt; Re(z) > 0; 0 B(z; u) = (z) (u)= (z + u); Pge 16, August 18, 1993

17 2F 1 (; b; c; z) = (c) () (b) 1X j=0 ( + j) (b + j) z j (c + j) j! : Further, E 1 (v) = Z 1 e vt dt; Re(v) > 0 1 t is the Exponentil Integrl nd erfc(z) is the Complementry Error Function Z erfc(z) = p 2 1 e t2 dt: For detils, see [1]. Built-in functions re vilble in the symbolic mnipultors Mple nd Mthemtic for ll the entries in the tbles bove nd we hve used them to compute high precision zero-th moments, for rel nd complex v without diculty Tbles for the results of numericl tests. Liner fctor -log(mximum errors) IC MS v N Tble 7 Errors for Jcobi mesure = 1=3; = 1=7, n = 10 z Liner fctor -log(mximum errors) IC MS v N (1117) (9033) (84032) Tble 8 Errors for Lguerre mesure = 1=3, n = 10 Pge 17, August 18, 1993

18 Qudrtic fctor v = x + iy -log(mximum errors) IC MS x y N (18049) Tble 9 Errors for Jcobi mesure = 1=3; = 1=7; n = 10 Qudrtic fctor v = x + iy -log(mximum errors) IC MS x y N (2442) (4591) (41270) Tble 10 Errors for Lguerre mesure = 0; n = 10 Qudrtic fctor v = x + iy -log(mximum errors) IC MS x y N (4749) (17591) ( ) Tble 11 Errors for Hermite mesure = 1=3; n = 10 Pge 18, August 18, 1993

19 Qudrtic fctor v = x + iy -log(mximum errors) IC MS x y N Tble 12 Errors for Lguerre mesure, v moving prllel to x xis. = 0; n = 10 Qudrtic fctor v = x + iy -log(mximum errors) IC MS x y N Tble 13 Errors for Legendre mesure n = 50 Qudrtic fctor v = x + iy -log(mximum errors) IC MS x y N (2980) (5319) (43398) Tble 14 Errors for Lguerre mesure n = 50 Pge 19, August 18, 1993

20 REFERENCES [1] M. Abrmowitz nd I.A. Stegun. Hndbook of mthemticl functions. Dover Publictions, New York, [2] T.S. Chihr. An Introduction to orthogonl polynomils. Gordon nd Brech, New York, [3] S. Elhy, G.H. Golub, nd J. Kutsky. Jcobi mtrices for sums of weight functions. BIT, 32:143{166, [4] S. Elhy nd J. Kutsky. Generlized Kronrod Ptterson type imbedded qudrtures. Aplikce Mtemtiky, 37(2):81{103, [5] W. Gutschi. Miniml solutions of three-term recurrence reltions nd orthogonl polynomils. Mth. Comp., 36(154):547{554, [6] W. Gutschi. On generting orthogonl polynomils. SIAM J. Sci. Sttist. Comput., 3:289{ 317, [7] J. Kutsky nd S. Elhy. Guss qudrtures nd Jcobi mtrices for weight functions not of one sign. Mth. Comp., 43(168):543{550, [8] J. Kutsky nd G.H. Golub. On the clcultion of Jcobi mtrices. Liner Algebr Appl., 52/53:439{455, [9] R.A. Sck nd A.F. Donovn. An lgorithm for Gussin qudrture given modied moments. Numer. Mth., 18:465{478, 1971/72. [10] G. Szego. Orthogonl polynomils. AMS Colloquium Publictions 23. Americm Mthemticl Society, Providence, Rhode, RI, 4th edition, [11] V.B. Uvrov. The connection between systems of polynomils orthogonl with respect to dierent distribution functions. USSR Computtionl Mth. nd Mth. Phys., 9(6):25{36, Pge 20, August 18, 1993

Matrices, Moments and Quadrature, cont d

Matrices, Moments and Quadrature, cont d Jim Lmbers MAT 285 Summer Session 2015-16 Lecture 2 Notes Mtrices, Moments nd Qudrture, cont d We hve described how Jcobi mtrices cn be used to compute nodes nd weights for Gussin qudrture rules for generl

More information

Lecture 14: Quadrature

Lecture 14: Quadrature Lecture 14: Qudrture This lecture is concerned with the evlution of integrls fx)dx 1) over finite intervl [, b] The integrnd fx) is ssumed to be rel-vlues nd smooth The pproximtion of n integrl by numericl

More information

1 The Lagrange interpolation formula

1 The Lagrange interpolation formula Notes on Qudrture 1 The Lgrnge interpoltion formul We briefly recll the Lgrnge interpoltion formul. The strting point is collection of N + 1 rel points (x 0, y 0 ), (x 1, y 1 ),..., (x N, y N ), with x

More information

1. Gauss-Jacobi quadrature and Legendre polynomials. p(t)w(t)dt, p {p(x 0 ),...p(x n )} p(t)w(t)dt = w k p(x k ),

1. Gauss-Jacobi quadrature and Legendre polynomials. p(t)w(t)dt, p {p(x 0 ),...p(x n )} p(t)w(t)dt = w k p(x k ), 1. Guss-Jcobi qudrture nd Legendre polynomils Simpson s rule for evluting n integrl f(t)dt gives the correct nswer with error of bout O(n 4 ) (with constnt tht depends on f, in prticulr, it depends on

More information

Math 1B, lecture 4: Error bounds for numerical methods

Math 1B, lecture 4: Error bounds for numerical methods Mth B, lecture 4: Error bounds for numericl methods Nthn Pflueger 4 September 0 Introduction The five numericl methods descried in the previous lecture ll operte by the sme principle: they pproximte the

More information

Lecture 6: Singular Integrals, Open Quadrature rules, and Gauss Quadrature

Lecture 6: Singular Integrals, Open Quadrature rules, and Gauss Quadrature Lecture notes on Vritionl nd Approximte Methods in Applied Mthemtics - A Peirce UBC Lecture 6: Singulr Integrls, Open Qudrture rules, nd Guss Qudrture (Compiled 6 August 7) In this lecture we discuss the

More information

ODE: Existence and Uniqueness of a Solution

ODE: Existence and Uniqueness of a Solution Mth 22 Fll 213 Jerry Kzdn ODE: Existence nd Uniqueness of Solution The Fundmentl Theorem of Clculus tells us how to solve the ordinry differentil eqution (ODE) du = f(t) dt with initil condition u() =

More information

ODE: Existence and Uniqueness of a Solution

ODE: Existence and Uniqueness of a Solution Mth 22 Fll 213 Jerry Kzdn ODE: Existence nd Uniqueness of Solution The Fundmentl Theorem of Clculus tells us how to solve the ordinry dierentil eqution (ODE) du f(t) dt with initil condition u() : Just

More information

NUMERICAL INTEGRATION. The inverse process to differentiation in calculus is integration. Mathematically, integration is represented by.

NUMERICAL INTEGRATION. The inverse process to differentiation in calculus is integration. Mathematically, integration is represented by. NUMERICAL INTEGRATION 1 Introduction The inverse process to differentition in clculus is integrtion. Mthemticlly, integrtion is represented by f(x) dx which stnds for the integrl of the function f(x) with

More information

New Expansion and Infinite Series

New Expansion and Infinite Series Interntionl Mthemticl Forum, Vol. 9, 204, no. 22, 06-073 HIKARI Ltd, www.m-hikri.com http://dx.doi.org/0.2988/imf.204.4502 New Expnsion nd Infinite Series Diyun Zhng College of Computer Nnjing University

More information

Theoretical foundations of Gaussian quadrature

Theoretical foundations of Gaussian quadrature Theoreticl foundtions of Gussin qudrture 1 Inner product vector spce Definition 1. A vector spce (or liner spce) is set V = {u, v, w,...} in which the following two opertions re defined: (A) Addition of

More information

ARITHMETIC OPERATIONS. The real numbers have the following properties: a b c ab ac

ARITHMETIC OPERATIONS. The real numbers have the following properties: a b c ab ac REVIEW OF ALGEBRA Here we review the bsic rules nd procedures of lgebr tht you need to know in order to be successful in clculus. ARITHMETIC OPERATIONS The rel numbers hve the following properties: b b

More information

Module 11.4: nag quad util Numerical Integration Utilities. Contents

Module 11.4: nag quad util Numerical Integration Utilities. Contents Qudrture Module Contents Module 11.4: ng qud util Numericl Integrtion Utilities ng qud util provides utility procedures for computtion involving integrtion of functions, e.g., the computtion of the weights

More information

Properties of Integrals, Indefinite Integrals. Goals: Definition of the Definite Integral Integral Calculations using Antiderivatives

Properties of Integrals, Indefinite Integrals. Goals: Definition of the Definite Integral Integral Calculations using Antiderivatives Block #6: Properties of Integrls, Indefinite Integrls Gols: Definition of the Definite Integrl Integrl Clcultions using Antiderivtives Properties of Integrls The Indefinite Integrl 1 Riemnn Sums - 1 Riemnn

More information

Math& 152 Section Integration by Parts

Math& 152 Section Integration by Parts Mth& 5 Section 7. - Integrtion by Prts Integrtion by prts is rule tht trnsforms the integrl of the product of two functions into other (idelly simpler) integrls. Recll from Clculus I tht given two differentible

More information

SUMMER KNOWHOW STUDY AND LEARNING CENTRE

SUMMER KNOWHOW STUDY AND LEARNING CENTRE SUMMER KNOWHOW STUDY AND LEARNING CENTRE Indices & Logrithms 2 Contents Indices.2 Frctionl Indices.4 Logrithms 6 Exponentil equtions. Simplifying Surds 13 Opertions on Surds..16 Scientific Nottion..18

More information

Orthogonal Polynomials

Orthogonal Polynomials Mth 4401 Gussin Qudrture Pge 1 Orthogonl Polynomils Orthogonl polynomils rise from series solutions to differentil equtions, lthough they cn be rrived t in vriety of different mnners. Orthogonl polynomils

More information

Math 520 Final Exam Topic Outline Sections 1 3 (Xiao/Dumas/Liaw) Spring 2008

Math 520 Final Exam Topic Outline Sections 1 3 (Xiao/Dumas/Liaw) Spring 2008 Mth 520 Finl Exm Topic Outline Sections 1 3 (Xio/Dums/Liw) Spring 2008 The finl exm will be held on Tuesdy, My 13, 2-5pm in 117 McMilln Wht will be covered The finl exm will cover the mteril from ll of

More information

Numerical quadrature based on interpolating functions: A MATLAB implementation

Numerical quadrature based on interpolating functions: A MATLAB implementation SEMINAR REPORT Numericl qudrture bsed on interpolting functions: A MATLAB implementtion by Venkt Ayylsomyjul A seminr report submitted in prtil fulfillment for the degree of Mster of Science (M.Sc) in

More information

Chapter 3 Polynomials

Chapter 3 Polynomials Dr M DRAIEF As described in the introduction of Chpter 1, pplictions of solving liner equtions rise in number of different settings In prticulr, we will in this chpter focus on the problem of modelling

More information

Here we study square linear systems and properties of their coefficient matrices as they relate to the solution set of the linear system.

Here we study square linear systems and properties of their coefficient matrices as they relate to the solution set of the linear system. Section 24 Nonsingulr Liner Systems Here we study squre liner systems nd properties of their coefficient mtrices s they relte to the solution set of the liner system Let A be n n Then we know from previous

More information

Numerical Integration

Numerical Integration Chpter 5 Numericl Integrtion Numericl integrtion is the study of how the numericl vlue of n integrl cn be found. Methods of function pproximtion discussed in Chpter??, i.e., function pproximtion vi the

More information

Unit #9 : Definite Integral Properties; Fundamental Theorem of Calculus

Unit #9 : Definite Integral Properties; Fundamental Theorem of Calculus Unit #9 : Definite Integrl Properties; Fundmentl Theorem of Clculus Gols: Identify properties of definite integrls Define odd nd even functions, nd reltionship to integrl vlues Introduce the Fundmentl

More information

How do we solve these things, especially when they get complicated? How do we know when a system has a solution, and when is it unique?

How do we solve these things, especially when they get complicated? How do we know when a system has a solution, and when is it unique? XII. LINEAR ALGEBRA: SOLVING SYSTEMS OF EQUATIONS Tody we re going to tlk bout solving systems of liner equtions. These re problems tht give couple of equtions with couple of unknowns, like: 6 2 3 7 4

More information

p-adic Egyptian Fractions

p-adic Egyptian Fractions p-adic Egyptin Frctions Contents 1 Introduction 1 2 Trditionl Egyptin Frctions nd Greedy Algorithm 2 3 Set-up 3 4 p-greedy Algorithm 5 5 p-egyptin Trditionl 10 6 Conclusion 1 Introduction An Egyptin frction

More information

Module 6: LINEAR TRANSFORMATIONS

Module 6: LINEAR TRANSFORMATIONS Module 6: LINEAR TRANSFORMATIONS. Trnsformtions nd mtrices Trnsformtions re generliztions of functions. A vector x in some set S n is mpped into m nother vector y T( x). A trnsformtion is liner if, for

More information

Numerical integration

Numerical integration 2 Numericl integrtion This is pge i Printer: Opque this 2. Introduction Numericl integrtion is problem tht is prt of mny problems in the economics nd econometrics literture. The orgniztion of this chpter

More information

Numerical Integration. 1 Introduction. 2 Midpoint Rule, Trapezoid Rule, Simpson Rule. AMSC/CMSC 460/466 T. von Petersdorff 1

Numerical Integration. 1 Introduction. 2 Midpoint Rule, Trapezoid Rule, Simpson Rule. AMSC/CMSC 460/466 T. von Petersdorff 1 AMSC/CMSC 46/466 T. von Petersdorff 1 umericl Integrtion 1 Introduction We wnt to pproximte the integrl I := f xdx where we re given, b nd the function f s subroutine. We evlute f t points x 1,...,x n

More information

Discrete Least-squares Approximations

Discrete Least-squares Approximations Discrete Lest-squres Approximtions Given set of dt points (x, y ), (x, y ),, (x m, y m ), norml nd useful prctice in mny pplictions in sttistics, engineering nd other pplied sciences is to construct curve

More information

Orthogonal Polynomials and Least-Squares Approximations to Functions

Orthogonal Polynomials and Least-Squares Approximations to Functions Chpter Orthogonl Polynomils nd Lest-Squres Approximtions to Functions **4/5/3 ET. Discrete Lest-Squres Approximtions Given set of dt points (x,y ), (x,y ),..., (x m,y m ), norml nd useful prctice in mny

More information

7.2 The Definite Integral

7.2 The Definite Integral 7.2 The Definite Integrl the definite integrl In the previous section, it ws found tht if function f is continuous nd nonnegtive, then the re under the grph of f on [, b] is given by F (b) F (), where

More information

Goals: Determine how to calculate the area described by a function. Define the definite integral. Explore the relationship between the definite

Goals: Determine how to calculate the area described by a function. Define the definite integral. Explore the relationship between the definite Unit #8 : The Integrl Gols: Determine how to clculte the re described by function. Define the definite integrl. Eplore the reltionship between the definite integrl nd re. Eplore wys to estimte the definite

More information

P 3 (x) = f(0) + f (0)x + f (0) 2. x 2 + f (0) . In the problem set, you are asked to show, in general, the n th order term is a n = f (n) (0)

P 3 (x) = f(0) + f (0)x + f (0) 2. x 2 + f (0) . In the problem set, you are asked to show, in general, the n th order term is a n = f (n) (0) 1 Tylor polynomils In Section 3.5, we discussed how to pproximte function f(x) round point in terms of its first derivtive f (x) evluted t, tht is using the liner pproximtion f() + f ()(x ). We clled this

More information

The Regulated and Riemann Integrals

The Regulated and Riemann Integrals Chpter 1 The Regulted nd Riemnn Integrls 1.1 Introduction We will consider severl different pproches to defining the definite integrl f(x) dx of function f(x). These definitions will ll ssign the sme vlue

More information

Math 270A: Numerical Linear Algebra

Math 270A: Numerical Linear Algebra Mth 70A: Numericl Liner Algebr Instructor: Michel Holst Fll Qurter 014 Homework Assignment #3 Due Give to TA t lest few dys before finl if you wnt feedbck. Exercise 3.1. (The Bsic Liner Method for Liner

More information

Physics 116C Solution of inhomogeneous ordinary differential equations using Green s functions

Physics 116C Solution of inhomogeneous ordinary differential equations using Green s functions Physics 6C Solution of inhomogeneous ordinry differentil equtions using Green s functions Peter Young November 5, 29 Homogeneous Equtions We hve studied, especilly in long HW problem, second order liner

More information

A REVIEW OF CALCULUS CONCEPTS FOR JDEP 384H. Thomas Shores Department of Mathematics University of Nebraska Spring 2007

A REVIEW OF CALCULUS CONCEPTS FOR JDEP 384H. Thomas Shores Department of Mathematics University of Nebraska Spring 2007 A REVIEW OF CALCULUS CONCEPTS FOR JDEP 384H Thoms Shores Deprtment of Mthemtics University of Nebrsk Spring 2007 Contents Rtes of Chnge nd Derivtives 1 Dierentils 4 Are nd Integrls 5 Multivrite Clculus

More information

AQA Further Pure 1. Complex Numbers. Section 1: Introduction to Complex Numbers. The number system

AQA Further Pure 1. Complex Numbers. Section 1: Introduction to Complex Numbers. The number system Complex Numbers Section 1: Introduction to Complex Numbers Notes nd Exmples These notes contin subsections on The number system Adding nd subtrcting complex numbers Multiplying complex numbers Complex

More information

Natural examples of rings are the ring of integers, a ring of polynomials in one variable, the ring

Natural examples of rings are the ring of integers, a ring of polynomials in one variable, the ring More generlly, we define ring to be non-empty set R hving two binry opertions (we ll think of these s ddition nd multipliction) which is n Abelin group under + (we ll denote the dditive identity by 0),

More information

Best Approximation in the 2-norm

Best Approximation in the 2-norm Jim Lmbers MAT 77 Fll Semester 1-11 Lecture 1 Notes These notes correspond to Sections 9. nd 9.3 in the text. Best Approximtion in the -norm Suppose tht we wish to obtin function f n (x) tht is liner combintion

More information

5.7 Improper Integrals

5.7 Improper Integrals 458 pplictions of definite integrls 5.7 Improper Integrls In Section 5.4, we computed the work required to lift pylod of mss m from the surfce of moon of mss nd rdius R to height H bove the surfce of the

More information

Review of Calculus, cont d

Review of Calculus, cont d Jim Lmbers MAT 460 Fll Semester 2009-10 Lecture 3 Notes These notes correspond to Section 1.1 in the text. Review of Clculus, cont d Riemnn Sums nd the Definite Integrl There re mny cses in which some

More information

Contents. Outline. Structured Rank Matrices Lecture 2: The theorem Proofs Examples related to structured ranks References. Structure Transport

Contents. Outline. Structured Rank Matrices Lecture 2: The theorem Proofs Examples related to structured ranks References. Structure Transport Contents Structured Rnk Mtrices Lecture 2: Mrc Vn Brel nd Rf Vndebril Dept. of Computer Science, K.U.Leuven, Belgium Chemnitz, Germny, 26-30 September 2011 1 Exmples relted to structured rnks 2 2 / 26

More information

Lecture 20: Numerical Integration III

Lecture 20: Numerical Integration III cs4: introduction to numericl nlysis /8/0 Lecture 0: Numericl Integrtion III Instructor: Professor Amos Ron Scribes: Mrk Cowlishw, Yunpeng Li, Nthnel Fillmore For the lst few lectures we hve discussed

More information

3.4 Numerical integration

3.4 Numerical integration 3.4. Numericl integrtion 63 3.4 Numericl integrtion In mny economic pplictions it is necessry to compute the definite integrl of relvlued function f with respect to "weight" function w over n intervl [,

More information

37 Kragujevac J. Math. 23 (2001) A NOTE ON DENSITY OF THE ZEROS OF ff-orthogonal POLYNOMIALS Gradimir V. Milovanović a and Miodrag M. Spalević

37 Kragujevac J. Math. 23 (2001) A NOTE ON DENSITY OF THE ZEROS OF ff-orthogonal POLYNOMIALS Gradimir V. Milovanović a and Miodrag M. Spalević 37 Krgujevc J. Mth. 23 (2001) 37 43. A NOTE ON DENSITY OF THE ZEROS OF ff-orthogonal POLYNOMIALS Grdimir V. Milovnović nd Miodrg M. Splević b Fculty of Electronic Engineering, Deprtment of Mthemtics, University

More information

8 Laplace s Method and Local Limit Theorems

8 Laplace s Method and Local Limit Theorems 8 Lplce s Method nd Locl Limit Theorems 8. Fourier Anlysis in Higher DImensions Most of the theorems of Fourier nlysis tht we hve proved hve nturl generliztions to higher dimensions, nd these cn be proved

More information

An approximation to the arithmetic-geometric mean. G.J.O. Jameson, Math. Gazette 98 (2014), 85 95

An approximation to the arithmetic-geometric mean. G.J.O. Jameson, Math. Gazette 98 (2014), 85 95 An pproximtion to the rithmetic-geometric men G.J.O. Jmeson, Mth. Gzette 98 (4), 85 95 Given positive numbers > b, consider the itertion given by =, b = b nd n+ = ( n + b n ), b n+ = ( n b n ) /. At ech

More information

Math 113 Exam 2 Practice

Math 113 Exam 2 Practice Mth 3 Exm Prctice Februry 8, 03 Exm will cover 7.4, 7.5, 7.7, 7.8, 8.-3 nd 8.5. Plese note tht integrtion skills lerned in erlier sections will still be needed for the mteril in 7.5, 7.8 nd chpter 8. This

More information

CMDA 4604: Intermediate Topics in Mathematical Modeling Lecture 19: Interpolation and Quadrature

CMDA 4604: Intermediate Topics in Mathematical Modeling Lecture 19: Interpolation and Quadrature CMDA 4604: Intermedite Topics in Mthemticl Modeling Lecture 19: Interpoltion nd Qudrture In this lecture we mke brief diversion into the res of interpoltion nd qudrture. Given function f C[, b], we sy

More information

Lesson 25: Adding and Subtracting Rational Expressions

Lesson 25: Adding and Subtracting Rational Expressions Lesson 2: Adding nd Subtrcting Rtionl Expressions Student Outcomes Students perform ddition nd subtrction of rtionl expressions. Lesson Notes This lesson reviews ddition nd subtrction of frctions using

More information

Advanced Calculus: MATH 410 Notes on Integrals and Integrability Professor David Levermore 17 October 2004

Advanced Calculus: MATH 410 Notes on Integrals and Integrability Professor David Levermore 17 October 2004 Advnced Clculus: MATH 410 Notes on Integrls nd Integrbility Professor Dvid Levermore 17 October 2004 1. Definite Integrls In this section we revisit the definite integrl tht you were introduced to when

More information

Lecture Note 9: Orthogonal Reduction

Lecture Note 9: Orthogonal Reduction MATH : Computtionl Methods of Liner Algebr 1 The Row Echelon Form Lecture Note 9: Orthogonl Reduction Our trget is to solve the norml eution: Xinyi Zeng Deprtment of Mthemticl Sciences, UTEP A t Ax = A

More information

UNIT 1 FUNCTIONS AND THEIR INVERSES Lesson 1.4: Logarithmic Functions as Inverses Instruction

UNIT 1 FUNCTIONS AND THEIR INVERSES Lesson 1.4: Logarithmic Functions as Inverses Instruction Lesson : Logrithmic Functions s Inverses Prerequisite Skills This lesson requires the use of the following skills: determining the dependent nd independent vribles in n exponentil function bsed on dt from

More information

State space systems analysis (continued) Stability. A. Definitions A system is said to be Asymptotically Stable (AS) when it satisfies

State space systems analysis (continued) Stability. A. Definitions A system is said to be Asymptotically Stable (AS) when it satisfies Stte spce systems nlysis (continued) Stbility A. Definitions A system is sid to be Asymptoticlly Stble (AS) when it stisfies ut () = 0, t > 0 lim xt () 0. t A system is AS if nd only if the impulse response

More information

Elements of Matrix Algebra

Elements of Matrix Algebra Elements of Mtrix Algebr Klus Neusser Kurt Schmidheiny September 30, 2015 Contents 1 Definitions 2 2 Mtrix opertions 3 3 Rnk of Mtrix 5 4 Specil Functions of Qudrtic Mtrices 6 4.1 Trce of Mtrix.........................

More information

Lecture 1. Functional series. Pointwise and uniform convergence.

Lecture 1. Functional series. Pointwise and uniform convergence. 1 Introduction. Lecture 1. Functionl series. Pointwise nd uniform convergence. In this course we study mongst other things Fourier series. The Fourier series for periodic function f(x) with period 2π is

More information

W. We shall do so one by one, starting with I 1, and we shall do it greedily, trying

W. We shall do so one by one, starting with I 1, and we shall do it greedily, trying Vitli covers 1 Definition. A Vitli cover of set E R is set V of closed intervls with positive length so tht, for every δ > 0 nd every x E, there is some I V with λ(i ) < δ nd x I. 2 Lemm (Vitli covering)

More information

Chapter 0. What is the Lebesgue integral about?

Chapter 0. What is the Lebesgue integral about? Chpter 0. Wht is the Lebesgue integrl bout? The pln is to hve tutoril sheet ech week, most often on Fridy, (to be done during the clss) where you will try to get used to the ides introduced in the previous

More information

Math 8 Winter 2015 Applications of Integration

Math 8 Winter 2015 Applications of Integration Mth 8 Winter 205 Applictions of Integrtion Here re few importnt pplictions of integrtion. The pplictions you my see on n exm in this course include only the Net Chnge Theorem (which is relly just the Fundmentl

More information

Abstract inner product spaces

Abstract inner product spaces WEEK 4 Abstrct inner product spces Definition An inner product spce is vector spce V over the rel field R equipped with rule for multiplying vectors, such tht the product of two vectors is sclr, nd the

More information

Introduction to Determinants. Remarks. Remarks. The determinant applies in the case of square matrices

Introduction to Determinants. Remarks. Remarks. The determinant applies in the case of square matrices Introduction to Determinnts Remrks The determinnt pplies in the cse of squre mtrices squre mtrix is nonsingulr if nd only if its determinnt not zero, hence the term determinnt Nonsingulr mtrices re sometimes

More information

Math 113 Exam 2 Practice

Math 113 Exam 2 Practice Mth Em Prctice Februry, 8 Em will cover sections 6.5, 7.-7.5 nd 7.8. This sheet hs three sections. The first section will remind you bout techniques nd formuls tht you should know. The second gives number

More information

Lecture Notes: Orthogonal Polynomials, Gaussian Quadrature, and Integral Equations

Lecture Notes: Orthogonal Polynomials, Gaussian Quadrature, and Integral Equations 18330 Lecture Notes: Orthogonl Polynomils, Gussin Qudrture, nd Integrl Equtions Homer Reid My 1, 2014 In the previous set of notes we rrived t the definition of Chebyshev polynomils T n (x) vi the following

More information

New data structures to reduce data size and search time

New data structures to reduce data size and search time New dt structures to reduce dt size nd serch time Tsuneo Kuwbr Deprtment of Informtion Sciences, Fculty of Science, Kngw University, Hirtsuk-shi, Jpn FIT2018 1D-1, No2, pp1-4 Copyright (c)2018 by The Institute

More information

f(x) dx, If one of these two conditions is not met, we call the integral improper. Our usual definition for the value for the definite integral

f(x) dx, If one of these two conditions is not met, we call the integral improper. Our usual definition for the value for the definite integral Improper Integrls Every time tht we hve evluted definite integrl such s f(x) dx, we hve mde two implicit ssumptions bout the integrl:. The intervl [, b] is finite, nd. f(x) is continuous on [, b]. If one

More information

Applicable Analysis and Discrete Mathematics available online at

Applicable Analysis and Discrete Mathematics available online at Applicble Anlysis nd Discrete Mthemtics vilble online t http://pefmth.etf.rs Appl. Anl. Discrete Mth. 4 (2010), 23 31. doi:10.2298/aadm100201012k NUMERICAL ANALYSIS MEETS NUMBER THEORY: USING ROOTFINDING

More information

20 MATHEMATICS POLYNOMIALS

20 MATHEMATICS POLYNOMIALS 0 MATHEMATICS POLYNOMIALS.1 Introduction In Clss IX, you hve studied polynomils in one vrible nd their degrees. Recll tht if p(x) is polynomil in x, the highest power of x in p(x) is clled the degree of

More information

THE EXISTENCE-UNIQUENESS THEOREM FOR FIRST-ORDER DIFFERENTIAL EQUATIONS.

THE EXISTENCE-UNIQUENESS THEOREM FOR FIRST-ORDER DIFFERENTIAL EQUATIONS. THE EXISTENCE-UNIQUENESS THEOREM FOR FIRST-ORDER DIFFERENTIAL EQUATIONS RADON ROSBOROUGH https://intuitiveexplntionscom/picrd-lindelof-theorem/ This document is proof of the existence-uniqueness theorem

More information

Improper Integrals. Type I Improper Integrals How do we evaluate an integral such as

Improper Integrals. Type I Improper Integrals How do we evaluate an integral such as Improper Integrls Two different types of integrls cn qulify s improper. The first type of improper integrl (which we will refer to s Type I) involves evluting n integrl over n infinite region. In the grph

More information

Chapter 7 Notes, Stewart 8e. 7.1 Integration by Parts Trigonometric Integrals Evaluating sin m x cos n (x) dx...

Chapter 7 Notes, Stewart 8e. 7.1 Integration by Parts Trigonometric Integrals Evaluating sin m x cos n (x) dx... Contents 7.1 Integrtion by Prts................................... 2 7.2 Trigonometric Integrls.................................. 8 7.2.1 Evluting sin m x cos n (x)......................... 8 7.2.2 Evluting

More information

How do we solve these things, especially when they get complicated? How do we know when a system has a solution, and when is it unique?

How do we solve these things, especially when they get complicated? How do we know when a system has a solution, and when is it unique? XII. LINEAR ALGEBRA: SOLVING SYSTEMS OF EQUATIONS Tody we re going to tlk out solving systems of liner equtions. These re prolems tht give couple of equtions with couple of unknowns, like: 6= x + x 7=

More information

Numerical Analysis. 10th ed. R L Burden, J D Faires, and A M Burden

Numerical Analysis. 10th ed. R L Burden, J D Faires, and A M Burden Numericl Anlysis 10th ed R L Burden, J D Fires, nd A M Burden Bemer Presenttion Slides Prepred by Dr. Annette M. Burden Youngstown Stte University July 9, 2015 Chpter 4.1: Numericl Differentition 1 Three-Point

More information

Riemann Sums and Riemann Integrals

Riemann Sums and Riemann Integrals Riemnn Sums nd Riemnn Integrls Jmes K. Peterson Deprtment of Biologicl Sciences nd Deprtment of Mthemticl Sciences Clemson University August 26, 203 Outline Riemnn Sums Riemnn Integrls Properties Abstrct

More information

Numerical Integration

Numerical Integration Chpter 1 Numericl Integrtion Numericl differentition methods compute pproximtions to the derivtive of function from known vlues of the function. Numericl integrtion uses the sme informtion to compute numericl

More information

Chapter 4 Contravariance, Covariance, and Spacetime Diagrams

Chapter 4 Contravariance, Covariance, and Spacetime Diagrams Chpter 4 Contrvrince, Covrince, nd Spcetime Digrms 4. The Components of Vector in Skewed Coordintes We hve seen in Chpter 3; figure 3.9, tht in order to show inertil motion tht is consistent with the Lorentz

More information

Improper Integrals, and Differential Equations

Improper Integrals, and Differential Equations Improper Integrls, nd Differentil Equtions October 22, 204 5.3 Improper Integrls Previously, we discussed how integrls correspond to res. More specificlly, we sid tht for function f(x), the region creted

More information

CLOSED EXPRESSIONS FOR COEFFICIENTS IN WEIGHTED NEWTON-COTES QUADRATURES

CLOSED EXPRESSIONS FOR COEFFICIENTS IN WEIGHTED NEWTON-COTES QUADRATURES Filomt 27:4 (2013) 649 658 DOI 10.2298/FIL1304649M Published by Fculty of Sciences nd Mthemtics University of Niš Serbi Avilble t: http://www.pmf.ni.c.rs/filomt CLOSED EXPRESSIONS FOR COEFFICIENTS IN WEIGHTED

More information

Before we can begin Ch. 3 on Radicals, we need to be familiar with perfect squares, cubes, etc. Try and do as many as you can without a calculator!!!

Before we can begin Ch. 3 on Radicals, we need to be familiar with perfect squares, cubes, etc. Try and do as many as you can without a calculator!!! Nme: Algebr II Honors Pre-Chpter Homework Before we cn begin Ch on Rdicls, we need to be fmilir with perfect squres, cubes, etc Try nd do s mny s you cn without clcultor!!! n The nth root of n n Be ble

More information

Riemann Sums and Riemann Integrals

Riemann Sums and Riemann Integrals Riemnn Sums nd Riemnn Integrls Jmes K. Peterson Deprtment of Biologicl Sciences nd Deprtment of Mthemticl Sciences Clemson University August 26, 2013 Outline 1 Riemnn Sums 2 Riemnn Integrls 3 Properties

More information

1 Linear Least Squares

1 Linear Least Squares Lest Squres Pge 1 1 Liner Lest Squres I will try to be consistent in nottion, with n being the number of dt points, nd m < n being the number of prmeters in model function. We re interested in solving

More information

Euler, Ioachimescu and the trapezium rule. G.J.O. Jameson (Math. Gazette 96 (2012), )

Euler, Ioachimescu and the trapezium rule. G.J.O. Jameson (Math. Gazette 96 (2012), ) Euler, Iochimescu nd the trpezium rule G.J.O. Jmeson (Mth. Gzette 96 (0), 36 4) The following results were estblished in recent Gzette rticle [, Theorems, 3, 4]. Given > 0 nd 0 < s

More information

NUMERICAL INTEGRATION

NUMERICAL INTEGRATION NUMERICAL INTEGRATION How do we evlute I = f (x) dx By the fundmentl theorem of clculus, if F (x) is n ntiderivtive of f (x), then I = f (x) dx = F (x) b = F (b) F () However, in prctice most integrls

More information

COSC 3361 Numerical Analysis I Numerical Integration and Differentiation (III) - Gauss Quadrature and Adaptive Quadrature

COSC 3361 Numerical Analysis I Numerical Integration and Differentiation (III) - Gauss Quadrature and Adaptive Quadrature COSC 336 Numericl Anlysis I Numericl Integrtion nd Dierentition III - Guss Qudrture nd Adptive Qudrture Edgr Griel Fll 5 COSC 336 Numericl Anlysis I Edgr Griel Summry o the lst lecture I For pproximting

More information

STEP FUNCTIONS, DELTA FUNCTIONS, AND THE VARIATION OF PARAMETERS FORMULA. 0 if t < 0, 1 if t > 0.

STEP FUNCTIONS, DELTA FUNCTIONS, AND THE VARIATION OF PARAMETERS FORMULA. 0 if t < 0, 1 if t > 0. STEP FUNCTIONS, DELTA FUNCTIONS, AND THE VARIATION OF PARAMETERS FORMULA STEPHEN SCHECTER. The unit step function nd piecewise continuous functions The Heviside unit step function u(t) is given by if t

More information

Quadratic Forms. Quadratic Forms

Quadratic Forms. Quadratic Forms Qudrtic Forms Recll the Simon & Blume excerpt from n erlier lecture which sid tht the min tsk of clculus is to pproximte nonliner functions with liner functions. It s ctully more ccurte to sy tht we pproximte

More information

Lecture 19: Continuous Least Squares Approximation

Lecture 19: Continuous Least Squares Approximation Lecture 19: Continuous Lest Squres Approximtion 33 Continuous lest squres pproximtion We begn 31 with the problem of pproximting some f C[, b] with polynomil p P n t the discrete points x, x 1,, x m for

More information

1.9 C 2 inner variations

1.9 C 2 inner variations 46 CHAPTER 1. INDIRECT METHODS 1.9 C 2 inner vritions So fr, we hve restricted ttention to liner vritions. These re vritions of the form vx; ǫ = ux + ǫφx where φ is in some liner perturbtion clss P, for

More information

Numerical Methods I Orthogonal Polynomials

Numerical Methods I Orthogonal Polynomials Numericl Methods I Orthogonl Polynomils Aleksndr Donev Cournt Institute, NYU 1 donev@cournt.nyu.edu 1 MATH-GA 2011.003 / CSCI-GA 2945.003, Fll 2014 Nov 6th, 2014 A. Donev (Cournt Institute) Lecture IX

More information

MATH 144: Business Calculus Final Review

MATH 144: Business Calculus Final Review MATH 144: Business Clculus Finl Review 1 Skills 1. Clculte severl limits. 2. Find verticl nd horizontl symptotes for given rtionl function. 3. Clculte derivtive by definition. 4. Clculte severl derivtives

More information

Recitation 3: More Applications of the Derivative

Recitation 3: More Applications of the Derivative Mth 1c TA: Pdric Brtlett Recittion 3: More Applictions of the Derivtive Week 3 Cltech 2012 1 Rndom Question Question 1 A grph consists of the following: A set V of vertices. A set E of edges where ech

More information

Construction of Gauss Quadrature Rules

Construction of Gauss Quadrature Rules Jim Lmbers MAT 772 Fll Semester 2010-11 Lecture 15 Notes These notes correspond to Sections 10.2 nd 10.3 in the text. Construction of Guss Qudrture Rules Previously, we lerned tht Newton-Cotes qudrture

More information

Review of basic calculus

Review of basic calculus Review of bsic clculus This brief review reclls some of the most importnt concepts, definitions, nd theorems from bsic clculus. It is not intended to tech bsic clculus from scrtch. If ny of the items below

More information

63. Representation of functions as power series Consider a power series. ( 1) n x 2n for all 1 < x < 1

63. Representation of functions as power series Consider a power series. ( 1) n x 2n for all 1 < x < 1 3 9. SEQUENCES AND SERIES 63. Representtion of functions s power series Consider power series x 2 + x 4 x 6 + x 8 + = ( ) n x 2n It is geometric series with q = x 2 nd therefore it converges for ll q =

More information

Chapter 14. Matrix Representations of Linear Transformations

Chapter 14. Matrix Representations of Linear Transformations Chpter 4 Mtrix Representtions of Liner Trnsformtions When considering the Het Stte Evolution, we found tht we could describe this process using multipliction by mtrix. This ws nice becuse computers cn

More information

Fourier Series and Their Applications

Fourier Series and Their Applications Fourier Series nd Their Applictions Rui iu My, 006 Abstrct Fourier series re of gret importnce in both theoreticl nd pplied mthemtics. For orthonorml fmilies of complex vlued functions {φ n}, Fourier Series

More information

Advanced Computational Fluid Dynamics AA215A Lecture 3 Polynomial Interpolation: Numerical Differentiation and Integration.

Advanced Computational Fluid Dynamics AA215A Lecture 3 Polynomial Interpolation: Numerical Differentiation and Integration. Advnced Computtionl Fluid Dynmics AA215A Lecture 3 Polynomil Interpoltion: Numericl Differentition nd Integrtion Antony Jmeson Winter Qurter, 2016, Stnford, CA Lst revised on Jnury 7, 2016 Contents 3 Polynomil

More information

Precalculus Spring 2017

Precalculus Spring 2017 Preclculus Spring 2017 Exm 3 Summry (Section 4.1 through 5.2, nd 9.4) Section P.5 Find domins of lgebric expressions Simplify rtionl expressions Add, subtrct, multiply, & divide rtionl expressions Simplify

More information

Section 6.1 INTRO to LAPLACE TRANSFORMS

Section 6.1 INTRO to LAPLACE TRANSFORMS Section 6. INTRO to LAPLACE TRANSFORMS Key terms: Improper Integrl; diverge, converge A A f(t)dt lim f(t)dt Piecewise Continuous Function; jump discontinuity Function of Exponentil Order Lplce Trnsform

More information

N 0 completions on partial matrices

N 0 completions on partial matrices N 0 completions on prtil mtrices C. Jordán C. Mendes Arújo Jun R. Torregros Instituto de Mtemátic Multidisciplinr / Centro de Mtemátic Universidd Politécnic de Vlenci / Universidde do Minho Cmino de Ver

More information