MIT Sloan School of Management

Size: px
Start display at page:

Download "MIT Sloan School of Management"

Transcription

1 MIT Sloan School of Managmnt Working Papr Rvisd: Jun 2003 Original: April 2001 SENSITIVITY ANALYSIS FOR SHORTEST PATH PROBLEMS AND MAXIMUM CAPACITY PATH PROBLEMS IN UNDIRECTED GRAPHS Ramkumar Ramaswamy, Jams B. Orlin, Nilopal Chakravarty 2003 by Ramkumar Ramaswamy, Jams B. Orlin, Nilopal Chakravarty. All rights rsrvd. Short sctions of txt, not to xcd two paragraphs, may b quotd without xplicit prmission, providd that full crdit including notic is givn to th sourc. This papr also can b downloadd without charg from th Social Scinc Rsarch Ntwork Elctronic Papr Collction:

2 Snsitivity Analysis for Shortst Path Problms and Maximum Capacity Path Problms in Undirctd Graphs Ramkumar Ramaswamy Infosys Tchnologis Limitd 3 rd Cross, Elctronics City Bangalor India ramkumarr@infy.com Jams B. Orlin MIT E Cambridg, MA jorlin@mit.du Nilopal Chakravarti Intgratd Dcision Systms Consultancy 1 Jalan Kilang Timor Singapor nilotpal@idsc.com.sg March 2001 Rvisd, Dcmbr 2003 Abstract. This papr addrsss snsitivity analysis qustions concrning th shortst path problm and th maximum capacity path problm in an undirctd ntwork. For both problms, w dtrmin th maximum and minimum wights that ach dg can hav so that a givn path rmains optimal. For both problms, w show how to dtrmin ths maximum and minimum valus for all dgs in O(m + K log K) tim, whr m is th numbr of dgs in th ntwork, and K is th numbr of dgs on th givn optimal path. Ky words: snsitivity analysis, shortst path problm, bottlnck shortst path, maximum capacity path problm. 1

3 1. Introduction. Lt G = (N, E) b an undirctd graph with n nods and m dgs, and dsignatd sourc nod s and sink nod t. This papr addrsss snsitivity analysis qustions concrning th shortst s-t path (SP) problm in G and th maximum capacity s-t path (MCP) problm in G. For ach (i, j) E, lt c ij dnot ithr th lngth of (i, j) or th capacity of (i, j) dpnding on whthr w ar solving th shortst path or th maximum capacity path problm. Suppos that P is a shortst s-t path in G. For ach dg E, th lowr SP tolranc of dg is th minimum non-ngativ valu that th lngth of dg can tak (with all othr lngths staying fixd) so that P rmains an optimal path. Similarly, th uppr SP tolranc of dg is th maximum valu in R that th lngth of dg can tak so that P rmains an optimal path. W show that th problm of finding all uppr and lowr tolrancs of dgs in E is computationally quivalnt to th Minimum Cost Intrval Problm which is as follows. For ach i = 1 to r, lt [a i, b i ] dnot an intrval with ndpoints in {1,..., q}, and lt d i b th associatd cost. For ach k = 1 to q, idntify a minimum cost intrval [a i, b i ] containing k. W provid an O(r + q log q) algorithm for th Minimum Cost Intrval Problm, and provid an O(m + P log P ) algorithm for finding all uppr and lowr tolrancs of dgs in E. Rlatd snsitivity analysis problms for tr solutions hav bn considrd initially by Shir and Witzgall (1980). Exampls of subsqunt rsarch in th sam dirction includ an O(mα (m,n)) 1 algorithm for snsitivity analysis of th minimum cost spanning tr problm by Tarjan (1984) and an O(m) algorithm for snsitivity analysis of minimum spanning trs and shortst path trs in planar graphs by Booth and Wstbrook (1994). W rfr th radr to Gal (1995), Grnbrg (1998) and to Gal and Grnbrg (1997) for xtnsiv rfrncs to a varity of snsitivity analysis problms in combinatorial optimization. W furthr considr th snsitivity analysis problm for th maximum capacity path problm. For a ntwork with dg capacitis, th capacity of a path is th minimum capacity of an dg on th path. Lt Q b th maximum capacity s-t path in G with rspct to capacity vctor c. For ach dg E, th lowr (rsp., uppr) MCP tolranc of th dg is th minimum (rsp., maximum) valu in R {-, }. that th capacity of dg can tak so that Q rmains a maximum capacity path. W show that th problm of finding all uppr and lowr tolrancs of dgs in E can b solvd in O(m + Q log Q ) tim. Morovr, th problm of finding all tolrancs is narly computationally quivalnt to th Minimum Cost Intrval Problm. Th most vital SP dg problm is th problm of finding an dg whos dltion from G incrass th lngth of th shortst path from s to t as much as possibl. Th most vital MCP dg problm is th problm of finding an dg whos dltion from G dcrass th maximum capacity of a path from s to t as much as possibl. Th most vital SP dg is th dg with th largst uppr tolranc, and th most vital MCP dg is th dg with th lowst MCP tolranc. Algorithms for th most vital dg problms on shortst paths includ Bar-Noy t al (1995), Malik t al (1989), Vnma t al (1996). Algorithms for finding th most vital dgs in minimum cost spanning trs hav bn dvlopd by Hsu t al (1991), Hsu t al. (1992), Iwano and Kato (1993), and Banrj and Saxna (1997). Th most vital dg problm for minimum cost spanning trs is not dirctly rlatd to th problm of idntifying dg tolrancs, although both ar fundamntal qustions of snsitivity analysis. 1 α(m, n) is th invrs Ackrmann function, and is VERY slowly growing in m and n. 2

4 A diffrnt approach for dtrmining uppr SP-tolrancs in O(m + n log n) tim for dgs in a shortst path P was indpndntly dvlopd by Hrshbrgr and Suri (2001). Thy incorrctly claimd to hav solvd th snsitivity analysis problm for dirctd graphs, and pointd out thir rror in Hrshbrgr and Suri (2002). This papr xtnds prvious rsults in th unpublishd work of Ramaswamy (1994), who gav O(n 2 ) algorithms for th uppr and lowr tolranc problms considrd hr. In Sctions 2, 3, and 4, w considr snsitivity analysis for th shortst path problm. In Sction 5, w addrss th Minimum Cost Intrval problm; w show its quivalnc to th snsitivity analysis problm for SP and valuat its computational complxity. In Sctions 6, and 7, w considr th maximum capacity path problm. W offr a summary and conclusions in Sction Uppr and Lowr Tolrancs for Combinatorial Optimization Problms. In this sction, w giv procdurs for calculating th uppr and lowr tolrancs of variabls for combinatorial optimization problms with linar costs as wll as for problms with bottlnck (or maximin) costs. W first considr linar objctivs. Combinatorial Optimization Problms with Linar Objctivs. Considr a combinatorial optimization problm X whr costs ar assumd to b non-ngativ. Lt P(c) dnot th instanc of X min (cx: x F), whr F {0, 1} n, and whr c 0. W assum that X is closd undr substitution of linar objctivs. That is, th instanc P(d) = min (dx: x F) is also an instanc of X so long as c and d hav th sam numbr of componnts and d 0. Lt z(c) dnot th optimal objctiv valu for P(c). Suppos that x is an optimal solution to P(c). W ar intrstd in how much th cost cofficints of c can chang with x rmaining optimal. To this nd, for ach indx i and for ach k R, w lt ik, c dnot th vctor drivd from c as follows:, ik, k if j = i; c j = c j, if j i.. For ach componnt i, w dfin th lowr tolranc α i := min { k: k 0 and x is optimal for, P( c ik )} to b th last non-ngativ valu that th cost of componnt i can tak so that x rmains optimal. Similarly, w dfin th uppr tolranc β i to b max { k: x ik, is optimal for P( c )}. If x is ik, optimal for P( c ) for all finit valus of k, thn βi :=. Th nxt thorm charactrizs th uppr and lowr tolrancs of x in trms of optimal solution valus for rlatd problms. Th thorm is asily stablishd, and w omit th proof. x i = 1, thn α i = 0, and i, 0 x = 0, thn β i =, and α = ( ). Thorm 1. Lt x b optimal for P(c), and lt i {1, 2,, n}. If β i, i,0 i = z( c ) c x. If i i cx z c 3

5 Th Shortst Path Problm W now rturn to th shortst path problm on an undirctd graph G = (N, E). Lt n= N and lt m E. For ach dg E, c dnots th lngth of dg. W assum that c 0 for all dgs. = W prmit costs to b infinit. If S is a subst of dgs, thn c (S) = S c. W lt G k, dnot th k, graph G = (N, E) in which th cost vctor c is rplacd by c. d k, Lt P dnot a shortst path from nod s to nod t in G, and lt d(s, t) = c( P ). In gnral, lt k,,0 ( s, t) b th lngth of th shortst s-t path in G. W not that if P, thn c ( P ) = c( P ) c. If w intrprt Thorm 1 in trms of tolrancs for th shortst path problm, w gt th following corollary. Corollary 1. Lt P b a shortst path in G = (N, E). If P, thn α = 0, and, β = d (,) s t c(p ) + c. If P, thn β = and α, = cp ( ) d 0 ( st, ). By Corollary 1, th lowr tolranc of an dg P is 0, and th uppr tolranc can b calculatd by solving a singl shortst path problm. Also by Corollary 1, th uppr tolranc of an dg P is infinity, and th lowr tolranc can b calculatd by solving a singl shortst path problm. Thus, w could find all of th dg tolrancs by solving at most m shortst path problms. In th nxt sctions, w will show how to find all dg tolrancs in O(m + P log P ) tim, which is narly a factor of m improvmnt in running tim. W will also show that th problm of computing all tolrancs rducs to th Minimum Cost Intrval Problm, as dfind in Sction 1. Combinatorial Optimization Problms with Bottlnck Objctivs. For a givn n-vctor c, and a 0-1 n-vctor x, lt cmin ( x) = min{ ci : xi = 1}. In this subsction, th vctor c dnots a vctor of capacitis, and c min (x) is th capacity of solution x. Considr th following instanc of a bottlnck combinatorial optimization problm Y: max (c min (x): x F), whr F {0, 1} n, and whr c j R {-, } for ach j = 1 to n. W assum Y is closd undr substitution of linar objctivs, and w lt B(c) dnot th instanc max (c min (x): x F). (Hr, w prmit all linar objctivs, with positiv and ngativ cofficints.) Lt v(c) dnot th optimal objctiv valu for B(c). Lt x b an optimal solution to B(c). Analogously to bfor, w lt th lowr tolranc α i and th uppr tolranc β i b dfind as follows: α i = min { k : x, is optimal for B ( c ik )}. β i = max { k : x, is optimal for B( c ik )}. If x, is optimal for B ( c ik ) for all ngativ valus of k, thn αi = -. If x, is optimal for B( c ik ) for all positiv valus of k, thn βi =. 4

6 Th following thorm is analogous to Thorm 1. It can b provd in a straightforward mannr, and w omit th proof. Thorm 2. Lt x b optimal for B(c), and lt i {1, 2,, n}. If 1. ( i, α i = vc ). If 2. If x is optimal for B ( c i, ), thn βi =. 3. If x is not optimal for B ( c i, i, ), thn βi = c ( min x ) x i = 0, thn th following ar tru: 4. α i =. i, 5. If vc ( ) = c ( x ), thn βi =. min i, 6. If vc ( ) c min ( x), thn β > i = c ( x ) min. x i = 1, thn th following ar tru: Th Maximum Capacity Path Problm W now rturn to th Maximum Capacity Path Problm. W lt from s to t in G. P dnot a maximum capacity path Th following corollary is a translation of Thorm 2 to th MCP problm. Corollary 2. Lt P b a maximum capacity path for G = (N, E). If P, thn th following ar tru:, 1. α = vg ( ) and 2. If P is a maximum capacity path for G,, thn β = ; 3. If P is not a maximum capacity path for G,, thn β is th minimum capacity of an dg of P \. If P, thn th following ar tru: 4. α = ; 5. If P is a maximum capacity path for G,, thn β = ; 6. If P is not a maximum capacity path for G,,thn, β = c ( P ) min. 3. Lowr S-P tolrancs. In this sction, w show how to dtrmin lowr S-P tolrancs for all dgs in O(m) tim. W first giv psudo-cod for th algorithm, and subsquntly stablish its corrctnss and running tim. Algorithm 1. Comput Lowr S-P Tolrancs. bgin for ach P, α := 0; dtrmin th shortst path lngth d(s, i) in G from s to i for all i N; dtrmin th shortst path lngth d(j, t) in G from j to t for all j N; for ach dg = (i, j) P, α = c(p ) - min (c(p ), d(s, i) + d(j, t), d(s, j) + d(i, t)); nd 5

7 Bfor stablishing th corrctnss of Algorithm 1, w introduc som mor notation. Lt T s dnot a tr of shortst paths from nod s to all othr nods, and lt T t b a tr of shortst paths to nod t. Lt P(s, i) dnot th path in T s from nod s to nod i. Lt P(j, t) dnot th path in T t from nod j to nod t. For any dg (i, j) E, lt W( i, j) = Psi (, ), ( ij, ), Pjt (, ), which is th s-t walk obtaind by concatnating P ( s, i), (i, j), and P( j,t). Thorm 3. Algorithm 1 corrctly computs th lowr S-P tolrancs for an undirctd ntwork G = (N, E), and can b implmntd to run in O(m) tim. Proof. By Corollary 1, it suffics to show that for dgs P, d,0 ( s, t ) = min (c(p ), d(s, i) + d(j, t), d(s, j) + d(i, t)), If thr is a shortst s-t path in G,0 that dos not contain dg, thn d,0 ( s, t ) = c(p ). If thr is a,0 shortst s-t path in G that dos contain dg, thn d,0 ( s, t) = min (d(s, i) + d(j, t), d(s, j) + d(i, t)). Computing d,0 (,) s t for all = (, i j) P rquirs only that w comput d( s, i) for all nods i and that w comput d( j,t) for all nods j. This rquirs only two shortst path computations on undirctd ntworks, which taks O(m) tim using th algorithm by Thorup (1997). 4. Uppr tolrancs and th Minimum Cost Intrval Problm In this sction, w giv an algorithm for computing uppr S-P tolrancs in G with rspct to a minimum lngth path P. W first giv psudo-cod for solving th uppr tolranc problm. W latr show that th bottlnck stp is quivalnt to th Minimum Cost Intrval Problm. Algorithm 2. Comput Uppr S-P Tolrancs. bgin for ach P, β := ; choos ε > 0 so that for any substs S and S of dgs, if c(s) < c(s ) thn c(s) + ε < c(s ); 2 for ach P, lt c = c + ε/n 2 ; for ach P, lt c = c + ε/n; lt T s b a tr of shortst paths from nod s to all othr nods with rspct to costs c ; lt T t b a tr of shortst paths to nod t from all othr nods with rspct to costs c ; for ach i N, lt P(s, i) dnot th path in T s from s to i; for ach j N, lt P(j, t) dnot th path in T t from j to t; for ach dg P, β := c c(p) + min (c(w(i, j): (i, j) E\P, and W(i, j)); nd W will soon stablish th corrctnss of Algorithm 2. Howvr, w first mak a brif commnt on th running tim. It might appar that thr ar four potntial bottlnck oprations. First of all, thr is th calculation of ε in th scond lin. Scond, thr is th calculation of th shortst path trs. Third, 2 Th purpos of ε is to prturb th problm so that shortst path P is uniqu and such that P T s T t 6

8 thr is th dfinition of W(i, j) for all (i,j) E\P. Fourth, thr is valuating for all in P th following: min{ c( W ( i, j)) : ( i, j) A, and W ( i, j)}. As for th calculation of ε, this can b accomplishd by choosing any positiv ε < 1 if all data ar intgral. If data is prmittd to b irrational, thn on can rprsnt th prturbation of costs implicitly as a scond componnt of th costs, and calculat th shortst path trs using lxicography. Th calculation of th shortst path trs using lxicography is O(m) using th tchniqu of Thorup (1997). W also do not dirctly dtrmin W(i, j). Instad, w show how to dtrmin for all in P min{ c( W(, i j)):(, i j) E\ P, and W(, i j)} by rducing it to th minimum cost intrval problm. This is accomplishd in Thorm 4 blow. Thus, it is th calculation of min{ c( W( i, j)) : ( i, j) E\ P, and W( i, j)} that is th bottlnck stp of Algorithm 2. W bgin proving th corrctnss of Algorithm 2 with th following lmma concrning th spanning trs T s and T t. Lmma 1. Suppos that T s, T t, P(s, j), and P(j, t) ar dfind as in Algorithm 2. Thn for ach dg P, P(s, j) P(j, t). Proof. W first not that for any two paths P and P in G, if c (P) c (P ), thn c(p) c(p ) by our choic of ε. So T s and T t ar shortst path trs with rspct to c. Morovr, P T s T t. Lt P(s, j) = P(s, i 1 ), Q 1, whr i 1 is th last nod of P on th path P(s, j). Lt P(j, t) = Q 2, P(i 2, t), whr i 2 is th first nod of P(j, t) on path P. Suppos P(s, i 1 ) P(i 2, t). In that cas, lt P (s, j) = P(s, i 2 ), Rv(Q 2 ), whr Rv(Q 2 ) is obtaind from Q 2 by visiting th dgs in opposit ordr. Similarly, lt P (j, t) = Rv(Q 1 ), P(i 1, t). Thn P (s, j) is an s-j path, P (j, t) is a j-t path, and c (P (s, j)) + c (P (j, t)) < c (P(s, j)) + c (P(j, t)), contradicting that P(s, j) and P(j, t) ar both shortst paths with rspct to c. Thorm 4. Algorithm 2 corrctly computs th uppr S-P tolrancs for an undirctd graph G = (N, E). Proof. By Corollary 1, for ach P, β =. So, w hncforth considr th cas that P. Also, by Corollary 1, it suffics to show th following: d, (,) s t = min { c( W( i, j)) : ( i, j) E\ P, and W( i, j)}, (1), If thr is no s-t path in G\, thn d (,) s t =, and th minimization in th right hand sid of (1) is ovr th null st, and so (1) is valid. W now considr th cas that P is som shortst s-t path in G\ with rspct to costs c. Sinc th lngth of th shortst s-t path is also th lngth of th shortst s-t walk, it follows that d, (,) s t min { c( W(, i j)):(, i j) E\ P, and W(, i j)} W nxt prov th rvrs inquality. 7

9 Lt S = { i N : P( s, i)}. Lt i dnot th last nod on P that is also in S, and lt j b th subsqunt nod on P. W know that nods i and j xist bcaus P is an s-t path, s S, and t S. By assumption, i S and j S, and so P(s, j). By Lmma 1, P(j, t). Bcaus P, it follows that (i, j). So, W(i, j). Path P is a concatnation of a path from nod s to nod i, dg (i, j), and, a path from nod j to nod t. Thrfor, c ( P ) = c( P ) d( s, i +c + d( j, t) = c( W( i, j)), complting th proof. ) ij Thorm 4 dos not gnraliz to dirctd graphs. W giv a countrxampl to th gnralization of Thorm 4 to dirctd graphs in Figur 1. It is th failur of Thorm 4 to gnraliz to dirctd graphs that, in our opinion, maks it mor difficult to comput dg tolrancs in dirctd graphs. s Figur 1. A countrxampl to th dirctd vrsion of Thorm 4, whr = (1, 2). 1 t Also, Thorm 4 would not b tru if dg costs could b 0 and if w did not rquir paths to b shortst paths with rspct to th prturbd costs c. Considr an undirctd vrsion of Figur 1 whr ach dg has a cost of 0. If w do not rquir paths to hav th fwst numbr of dg, w can lt ach of P(s, 1), P(s, 2), P(s, 3), P(1, t), P(2, t), and P(3, t) contain dg = (1, 2). In this cas, th minimization in Thorm 1 would b ovr th mpty st. W now us Thorm 4 to transform th problm of computing uppr tolrancs for dgs in to th Minimum Cost Intrval Problm. W accomplish this in Algorithm 3. Algorithm 3. Transform SP Uppr Tolranc Problm to Min Cost Intrval Problm bgin rlabl dgs so that P = 1, 2,, K ; lt T s, T t, P(s, j), P(j, t) b dfind as in Algorithm 2; d(s, j) := c(p(s, j)) for j = 1 to n; d(j, t) := c(p(j, t)) for j = 1 to n; for ach i N do bgin if P(s, i) P =, thn a(i) := 1; ls choos a(i) so that a(i)-1 is th last dg of P that is also on P(s, i); if P(i, t) P =, thn b(i) := K; ls choos b(i) so that b(i)+1 is th first dg of P that is also on P(i, t); nd for ach dg (i, j) P with a(i) b(j), crat an intrval [a(i), b(j)] with cost c(w(i, j)) = d(s, i) + c ij + d(j, t); nd Givn th shortst path trs from Algorithm 2, w can asily comput a(i) in Algorithm 3 for all nods i in O(n) tim by scanning nods of T s in dpth first sarch ordr. If j is th prdcssor of i in T s, P 8

10 and if j P, thn a(i) is th indx of th dg following j on P. If j P, thn a(i) = a(j). Similarly, w can comput b(i) for all nods i in O(n) tim by scanning nods of T t in dpth first sarch ordr. b ( j) + 1 Bcaus P T s T t, it follows that all dgs prcding on P(j, t) ar also on P. a ( i) 1 in P(s, i) and all dgs succding Thorm 5. For ach dg (i, j) P with a(i) b(j), crat an intrval [a(i), b(j)] with cost c(w(i, j)) = d(s, i) + c ij + d(j, t) as in Algorithm 3. Thn th minimum cost of an intrval covring k is d k, (,) s t = min { c( W(, i j)):(, i j) E\ P, and W(, i j)}., Proof. By Thorm 3 and Algorithm 2, d k (,) s t = min{ c( W(, i j)):(, i j) E \ P and k W(, i j)}. W will complt th proof by showing that for any dg k, w hav k W(i, j) if and only if k [a( i), b( j)]. If k W(i, j), thn k P(s, i) or k P(j, t) or both. Hnc k < a(i) or k > b(j) or both, and thus k [a(i), b(j)]. Convrsly, if k W(i, j), thn k P(s, i) and k P(j, t). Hnc k > a(i)-1 and k < b(j) + 1, and thus k [a(i), b(j)]. Lt TOL(K, m) dnot th tim to find th uppr and lowr tolrancs for a shortst path problm on an undirctd graph with m dgs and with K dgs in th givn shortst s-t path. Lt INT(q, r) dnot th tim to solv th Minimum Cost Intrval Problm ovr r intrvals with ndpoints in {1,..., q}. Th nxt thorm stats th computational quivalnc of th Minimum Cost Intrval Problm and th problm of finding uppr and lowr tolrancs for a shortst path problm on an undirctd graph. Thorm 6. Suppos that m n. Thn TOL(K, m) = O(INT(K, m)), and INT(q, r) = O(TOL(q, r)). Proof. Thorms 3 and 4 show that computing th tolrancs of dgs not on P rquirs th computation of two shortst path problms, and this taks O(m) tim (Thorup (1997)). Thorms 1 and 5 show that th additional tim to comput tolrancs of dgs on P is O(INT(K, m)). Thrfor TOL(K, m) = O(INT(K, m)). Now suppos that r q, and considr th Minimum Cost Intrval Problm in which th intrvals ar [a(i), b(i)], with cost d(i). Without loss of gnrality, assum that thr is at most on intrval with ndpoints i and j for all i and j. If thr is an intrval [j, j], lt z j dnot its cost. Othrwis, lt z j =. W crat a tolranc problm as follows. W crat a graph G = (N, E), with q + 1 nods, whr th sourc nod is nod 1, and th sink nod is nod q+1. Lt P b th path 1, 2,..., q+1, and suppos that ach dg of P has a cost of 0. Lt i dnot th dg (i, i+1). For ach i = 1 to r with a(i) b(i), thr is an dg r+i = (a(i), b(i)+1) with a cost of d(i). If an intrval i is such that a(i) = b(i), w call it a zro lngth intrval. W do not crat dgs in G corrsponding to zro-lngth intrvals. Lt β j dnot th uppr tolranc of dg j in G. Th minimum cost of an intrval covring j is min( di ( ) : j [ ai ( ), bi ( )]), which w will show is qual to min { βj, z j }. If j [a(i), b(i)] and a(i) b(i) thn j P( 1, a( i)), and P( b( i) + 1, q+ 1), and so W ( a( i), b( i) + 1). Similarly, if j [a(i), b(i)], thn j j ithr j < a(i), and j P( 1, a( i)) or ls j > b(i), and P( b( i) + 1, q+ 1). So, j is containd in a nonzro lngth intrval [a(i), b(i)] if and only if W ( a( i), b( i) + 1). By Thorm 5, th cost of a minimum cost j j 9

11 nonzro lngth intrval covring j is th uppr tolranc of j in G. This stablishs that INT(q, r) = O(TOL(q+1, q+r)). Howvr, sinc th tolranc problm can b solvd in polynomial tim and sinc q r by assumption, it follows that TOL(q+1, q+r) = O(TOL(q, r)), and accordingly INT(q, r) = O(TOL(q, r)). 5. Solving th Minimum Cost Intrval Problm. Hr w provid an O(r + q log q) algorithm for solving th Minimum Cost Intrval Problm. Shigno and Uno (2002) indpndntly solvd th Minimum Cost Intrval Problm with th sam running tim. Lt F dnot a st of r intrvals. Lt c ij dnot th cost of th intrval [i, j] for ach [i, j] F. Th minimum cost of an intrval covring intgr l is g(l) = min { c ij : i l j}. Lt f k (j) = min { c ij : 1 i k }. Thn g(k) = min { f k (j) : k j q}. Morovr, for ach k and for ach j, f k (j) = min {f k-1 (j), c kj }. Th following algorithm computs th minimum cost of ach intrval. Algorithm 4. Th Minimum Cost Intrval Algorithm bgin for j =1 to q, f(j) = min {c 1j, }; g(1) := min{ f(j): j = 1 to q}; for k = 2 to q do bgin for j = k to q do f(j) = min {f(j), c kj }; g(k) = min{ f(j): j = k to q}; nd nd Th first for loop computs f 1 (j) for ach j. Whn th indx is k, th scond for loop computs f k (j) for ach j. Th corrctnss of th algorithm follows from th fact that minimum cost of an intrval covring intgr k is min {f k (j): j = k to q}. Th algorithm can b implmntd in O(r + q log q) using Frdman and Tarjan s (1984) Fibonacci Hap. At ach itration, w stor th valus of f( ) in th hap. To initializ th hap taks O(q) stps. To updat th hap in th lin f(j) = min {f(j), c kj } taks O(r) stps in total sinc ach intrval causs f to ithr stay th sam or dcras, and th dcras opration taks O(1) stps. Finally, thr ar q oprations of finding th min and q options of dlting f(k) at th nd of itration k. Each of ths oprations taks O(log q) tim using Fibonacci Haps. W stat our conclusions in th nxt thorm. Thorm 7. A Fibonacci Hap implmntation of Algorithm 4 solvs th Minimum Cost Intrval Problm in O(r + q log q) stps. 6. Uppr Tolrancs for th Maximum Capacity Path Problm In this sction, w rturn to th Maximum Capacity Path Problm on a ntwork G = (N, E), whr c dnots th capacity of dg E. 10

12 Th MCP problm ariss in svral domains. For xampl, on mthod for implmnting th augmnting path algorithm for th maximum flow problm is to snd flow along a path with maximum capacity. This was first analyzd by Edmonds and Karp (1972). Additional dtails can b found in Ahuja t. al., (1993). Th maximum augmnting path problm is mathmatically quivalnt to th bottlnck shortst path problm. An xampl of th bottlnck shortst path problm is th problm of finding a path from s to t such that th minimum rliability of an dg is maximizd. Lt P dnot a maximum capacity path from s to t in G. W nxt us th rsults of Corollary 2 to provid algorithms for computing th uppr tolrancs of all dgs in O(m) tim. Algorithm 5. Comput Uppr MCP tolrancs for dgs not in P bgin H := {a E: c a > c min (P)}; lt G H dnot th graph (N, H); lt S b th nods in th sam connctd componnt as s in G H ; lt T b th nods in th sam connctd componnt as t in G H ; for ach dg = (i, j) E\P do if i S and j T or if i T and j S, thn β = c min (P); ls β = ; nd Algorithm 6. Comput Uppr MCP tolrancs for dgs in P. bgin for ach P such that c c min (P), β = ; lt b any minimum capacity dg in P; lt H := {a E: c a > c min (P\ ) }; lt G H dnot th graph (N, H); lt S b th nods in th sam connctd componnt as s in G H ; lt T b th nods in th sam connctd componnt as t in G H ; for ach = (i, j) P such that c = c min (P) do; bgin if i S and j T or if i T and j S, thn β = c min (P\ ); ls β = ; nd nd Thorm 8. Algorithm 5 corrctly computs th uppr MCP tolrancs for dgs not in P. Algorithm 6 corrctly computs th uppr MCP tolrancs for dgs in P. Each algorithms can b implmntd to run in O(m) tim. Proof. For ach dg = (i, j) P, Algorithm 5 dtrmins whthr thr is an s-t path in G, whos capacity xcds cmin(p ), and sts β corrctly using th rsults of Corollary 2. For ach dg = (i, j) P with c > c min (P ), Algorithm 6 sts β =, as pr Corollary 2. W now considr ach dg = (i, j) P with c = c min (P ). Th capacity of P in G, is cmin(p \), which is th scond smallst capacity of an dg of P. Algorithm 6 dtrmins if P is a maximum capacity path in G, by chcking whthr thr is som path with capacity gratr than cmin(p \), and thn sts sts β corrctly using th rsults of Corollary 2. For Algorithms 5 and 6, dtrmining H and th connctd componnts taks O(m) tim, as dos computing c min (P \), complting th proof. 11

13 7. Efficint Computation of Lowr Tolrancs for th MCP Problm In this sction, w comput lowr tolrancs for dgs P. Our rsults rly on a clos connction btwn maximum capacity paths and th maximum capacity spanning tr. Th maximum capacity spanning tr (MST) problm for G is to find a spanning tr T for which c is maximum. W will rfr to an optimum solution as a maximum capacity spanning tr. T Th following lmma is asily stablishd, and is wll known. Lmma 2. Lt T dnot a maximum capacity spanning tr of G = (N, E). For any pair of nods i and j, th path in T from i to j is a maximum capacity path from i to j in G. Hncforth, w lt T dnot som maximum capacity spanning tr of G. W lt b(s, j) dnot th capacity of th path in T from s to j, and w lt b(j, t) dnot th capacity of th path in T from j to t. Th tr T can b calculatd in linar tim via a randomizd algorithm as pr th tchniqu of (Kargr t. al., 1995). Th bst dtrministic algorithm for computing th minimum cost spanning tr is du to Gabow t. al., (1984), and th running tim is O(m f ( nm, )), whr f ( nm, ) = ( i) min( i : log n m / n). Th valus b(s, j) and b(j, t) can b computd for all j in an additional O(n) tim. Th notation b was slctd sinc th maximum capacity of a path is th capacity of its "bottlnck" dg. Th following algorithm computs lowr tolrancs in G. W will subsquntly prov its corrctnss and show how its running tim rducs to that of finding a maximum capacity spanning tr and solving an instanc of th Minimum Cost Intrval Problm. Algorithm 7. Comput Lowr MCP Tolrancs bgin for ach P, α = - ; lt T b a maximum capacity spanning tr; for ach dg P \T, α = c min (P ); lt P(i, j) dnot th path in T from nod i to nod j; lt b(i, j) = c min (P(i, j)) dnot th capacity of th path from i to j; for ach P T, α = max{ (min(b(s, i), c ij, b(j, t)) : (i, j), and T + (i, j) is a tr }; nd Thorm 9. Th algorithm Comput Lowr MCP Tolrancs corrctly computs th lowr tolrancs of capacitis in th ntwork G = (N, E), givn a maximum capacity s-t path P. Proof. For ach dg P, α = - by Corollary 2. Also by Corollary 2, for ach dg P, α is th maximum capacity of a path in G,-, which is dnotd as vc (, ). If P and T, thn by Lmma 2, P(s, t) is a maximum capacity path in G, and thus it is also a maximum capacity path in G,-. It follows that α = c min (P ). If P T and if thr is no path in G\ from s to t, thn Algorithm 7 corrctly sts α to - as thr is no arc (i, j) for which T + (i, j) is a tr. 12

14 Finally w considr dgs P T such that thr is a path in G\ from s to t. For ach dg (i, j) T, lt W(i, j) = P(s, i), (i, j), P(j, t), which is a walk from s to t containing dg (i, j). If W(i, j), contains dg, thn c ( min W (, )), i j =. Othrwis cmin ( W( i, j)) = min( b( s, i), cij, b( j, t)). Sinc W(i, j) contains a path from s to t whos capacity is at last as grat as vc (, ) min{b(s, i), c ij, b(j, t)) : W(i, j) dos not contain dg }. c ( W( i, j)) min, it follows that Morovr, W(i, j) dos not contain dg prcisly whn T + (i, j) is a tr. Thrfor, α max{ (min(b(s, i), c ij, b(j, t)) : T + (i, j) is a tr }. W now prov th rvrs inquality in th cas whn P T and thr is a path from s to t in G\. Lt P b a maximum capacity s-t path in G,-. Lt S dnot th nods in T \ in th sam componnt as s. Thn N\S ar th nods of T \ in th sam componnt as t. Thn P contains som dg (i, j) with i S, and j S. Also, P. So T + (i, j) is a tr. Morovr,, c ( P ) = c ( P ) min( b( s, i), c, b( j, t)), with th lattr inquality following from th fact that th path min min ij in P from s to i has capacity at most b(s, i) from Lmma 2, and th path in P from j to t has capacity at most b(j, t). W conclud that α = c min (P ) max{ (min(b(s, i), c ij, b(j, t)) : T + (i, j) is a tr }, and so th thorm is provd. W now us Thorm 9 to transform th problm of computing tolrancs for dgs in P to th Maximum Cost Intrval Problm. This is quivalnt to th Minimum Cost Intrval Problm xcpt that w want to dtrmin th maximum cost intrval containing k for ach k = 1 to q. Our transformation is vry similar to th on containd in Algorithm 3. Lt T b th maximum capacity spanning tr. For ach nod i N, rcall that P(s, i) is a maximum capacity path from s to i in T, and P(j, t) is a maximum capacity path from j to t in T. W assum that th dgs of E ar ordrd so that P(s, t) P consists of th dgs 1, 2,..., r in th ordr that thy appar on th path P(s, t). For ach nod i, if P has no dg in common with P(s, i), thn lt a(i) = 1. Othrwis, lt dnot th highst indx dg of P(s, t) P that is also on P(s, i). If P a ( i) 1 has no dg in common with P(j, t), thn lt b(j) = r. Othrwis, lt b ( j) + 1 dnot th last indx dg of P(s, t) P that is also on P(j, t). W can comput a(i) and b(i) for all i N in O(n) tim in a similar mannr to th way indics ar computd for Algorithm 3. Lmma 3. Considr th Maximum Cost Intrval Problm, dfind as follows: For ach dg (i, j) P, with a(i) b(j), thr is an intrval [a(i), b(j)] with cost minimum cost of an intrval covring k is c ( W( i, j)) min. Thn th k, vc ( )= max (c min (W(i, j) : k W(i, j) ). Proof. Essntially th sam as th proof of Thorm 4. 13

15 W lt TOLCAP(K, m) dnot th tim to find tolrancs for th maximum capacity path problm, whr K is th numbr of arcs of th givn path. Lt MST(n, m) b th tim to solv a minimum cost spanning tr problm on n nods and m dgs. 3 Thorm 10. Suppos that m n. Thn TOLCAP(K, m) = O(INT(K, m) + MST(n, m)). Morovr, INT(r, q) = O(TOLCAP(r, q)). Proof. Th proof that TOLCAP(K, m) = O(INT(K, m) + MST(n, m)) rlis on th fact that on can comput tolrancs by solving a maximum cost intrval problm and by finding a maximum cost spanning tr. Th othr dtails ar th sam as in th proof of Thorm 6. Now suppos that r q, and considr th Maximum Cost Intrval Problm in which th intrvals ar [a(i), b(i)], with cost d(i). Th proof rlis on th sam construction as in th proof of Thorm 6, xcpt that hr P is a path 1, 2,..., q + 1, whr ach dg of th path has a capacity of M > d max. 8. Summary and conclusion In this papr w hav considrd snsitivity analysis qustions for th shortst s-t path (SP) and maximum capacity s-t path (MCP) problms and prsntd algorithms for answring ths qustions that ar far suprior to succssiv roptimization. Tabl 1 summarizs our contribution. Tabl 1. Summary of rsults Problm Complxity 1 Minimum [Maximum] cost intrval INT(r, q) = O(r + q log q) Shortst path snsitivity 2 Lowr tolrancs of dgs O(m) 3 Uppr tolrancs of dgs O(INT(n, m)) = O(m + n log n) Maximum capacity path snsitivity 4 Lowr tolrancs of dgs O(MST(n, m) + INT(n, m)) = O(m + n log n) 5 Uppr tolrancs of dgs O(m) Som opn qustions includ th following. (1) What is th computational complxity for th snsitivity analysis qustions addrssd in this papr if on prmits ngativ cost dgs in th SP, but no ngativ cost cycl? (2), What is th computational complxity for th snsitivity analysis qustions addrssd in this papr if considrs dirctd rathr than undirctd graphs? (Som partial rsults hav rcntly bn obtaind by Hrshbrgr t al. (2003)) and (3) Ar thr suprior algorithms for solving th Minimum Cost Intrval Problm? Acknowldgmnts 3 Th maximum capacity spanning tr is mathmatically quivalnt to th minimum cost spanning tr problm, and so has th sam running tim. 14

16 This work was supportd in part by Offic of Naval Rsarch grant ONR N W thank Dan Stratila and two anonymous rfrs for thir numrous suggstions lading to improvmnts in xposition. Rfrncs Ahuja R., T. Magnanti and J. Orlin Ntwork Flows: Thory, Algorithms, and Applications, Prntic- Hall, Englwood Cliffs, NJ. Banrj, S, and S. Saxna Paralll algorithm for finding th most vital dg in wightd graphs. Journal of Paralll and Distributd Computing 46, Bar-Noy, A., S. Khullr, and B. Schibr Th complxity of finding most vital dgs and nods. Tchnical Rport CS-TR-3539, Univrsity of Maryland Institut for Advancd Studis, Collg Park. Booth H. and J. Wstbrook A linar algorithm for analysis of minimum spanning and shortst path trs of planar graphs. Algorithmica 11, Edmonds J. and R. M. Karp Thortical improvmnts in algorithmic fficincy for ntwork flow problms. Journal of th ACM, 19, Frdman, M.L., and R.E. Tarjan Fibonacci haps and thir uss in improvd ntwork optimization algorithms, Procdings of th 25th Annual IEEE Symposium on th Foundation of Computr Scinc, Full papr in Journal of th ACM 34 (1987) Gabow H. N., Z. Galil, and T.H. Spncr Efficint implmntations of graph algorithms using contraction. Procdings of th 25 th Annual Symposium on th Foundations of Computr Scinc (FOCS '84), Gal. T Snsitivity Analysis, Paramtric Programming, and Rlatd Topics: Dgnracy, Multicritria Dcision Making, Rdundancy W.dGruytr, Brlin and Nw York. Gal, T. and H. J. Grnbrg (ds.) Advancs in Snsitivity Analysis and Paramtric Programming. Volum 6 of Intrnational Sris in Oprations Rsarch and Managmnt Scinc, Kluwr Acadmic Publishrs, Boston. Grnbrg H. J An annotatd bibliography for post-solution analysis in mixd intgr programming and combinatorial optimization, in D. Woodruff (Ed.), Advancs in Computational and Stochastic Optimization, Logic Programming, and Huristic Sarch, Kluwr Acadmic Publishrs, Hrshbrgr, J. and S. Suri Vickry prics and shortst paths: What is an dg worth? In Proc. 42nd Annu. IEEE Symp. Found. Comput. Sci., pags , Hrshbrgr, J. and S. Suri Erratum to Vickry prics and shortst paths: What is an dg worth? Procdings of th 43 rd Annual IEEE Symposium on Foundations of Computr Scinc (FOCS 02), Hrshbrgr, J. and S. Suri, and A. Bhosl On th difficulty of som shortst path problms. Procdings of th Symposium on Thortical Aspcts of Computr Scinc (STACS 2003), Hsu, L. H., R.H. Jan, Y.C. L, C.N. Hung, and M.S. Chrn Finding th most vital dg with rspct to minimum spanning tr in wightd graphs. Information Procssing Lttrs 39,

17 Hsu, L.H., P.F. Wang, and C.T. Wu Paralll algorithms for finding th most vital dg with rspct to minimum spanning tr. Paralll Comput. 18, Kargr D. R., P. N. Klin, and R. E. Tarjan A randomizd linar-tim algorithm to find minimum spanning trs. Journal of th ACM, 42, Malik, K., A.K. Mittal, and S.K. Gupta Th k most vital dgs in th shortst path problm. Oprations Rsarch Lttrs 8, Ramaswamy R Snsitivity Analysis in Combinatorial Optimization. Unpublishd Fllow Programm dissrtation, Indian Institut of Managmnt Calcutta, India. Shir D. R. and C. Witzgall dg tolrancs in shortst path and ntwork flow problms. Ntworks 10, 277. Shigno, M, and T. Uno Prsonal Corrspondnc. Tarjan R.E Snsitivity Analysis of Minimum Spanning Trs and Shortst Path Trs. Information Procssing Lttrs 14, Tarjan R. E A simpl vrsion of Karzanov s blocking flow algorithm. Oprations Rsarch Lttrs 2, Thorup M Undirctd Singl Sourc Shortst Paths in Linar Tim. Procdings of th 38 th Annual Symposium on th Foundations of Computr Scinc (FOCS '97). Vnma, S., H. Shn and F. Surawra NC Algorithms for th Singl Most Vital Edg Problm with Rspct to Shortst Paths, Information Procssing Lttrs 60,

Derangements and Applications

Derangements and Applications 2 3 47 6 23 Journal of Intgr Squncs, Vol. 6 (2003), Articl 03..2 Drangmnts and Applications Mhdi Hassani Dpartmnt of Mathmatics Institut for Advancd Studis in Basic Scincs Zanjan, Iran mhassani@iasbs.ac.ir

More information

Basic Polyhedral theory

Basic Polyhedral theory Basic Polyhdral thory Th st P = { A b} is calld a polyhdron. Lmma 1. Eithr th systm A = b, b 0, 0 has a solution or thr is a vctorπ such that π A 0, πb < 0 Thr cass, if solution in top row dos not ist

More information

The Matrix Exponential

The Matrix Exponential Th Matrix Exponntial (with xrciss) by D. Klain Vrsion 207.0.05 Corrctions and commnts ar wlcom. Th Matrix Exponntial For ach n n complx matrix A, dfin th xponntial of A to b th matrix A A k I + A + k!

More information

Week 3: Connected Subgraphs

Week 3: Connected Subgraphs Wk 3: Connctd Subgraphs Sptmbr 19, 2016 1 Connctd Graphs Path, Distanc: A path from a vrtx x to a vrtx y in a graph G is rfrrd to an xy-path. Lt X, Y V (G). An (X, Y )-path is an xy-path with x X and y

More information

The Matrix Exponential

The Matrix Exponential Th Matrix Exponntial (with xrciss) by Dan Klain Vrsion 28928 Corrctions and commnts ar wlcom Th Matrix Exponntial For ach n n complx matrix A, dfin th xponntial of A to b th matrix () A A k I + A + k!

More information

Homework #3. 1 x. dx. It therefore follows that a sum of the

Homework #3. 1 x. dx. It therefore follows that a sum of the Danil Cannon CS 62 / Luan March 5, 2009 Homwork # 1. Th natural logarithm is dfind by ln n = n 1 dx. It thrfor follows that a sum of th 1 x sam addnd ovr th sam intrval should b both asymptotically uppr-

More information

Computing and Communications -- Network Coding

Computing and Communications -- Network Coding 89 90 98 00 Computing and Communications -- Ntwork Coding Dr. Zhiyong Chn Institut of Wirlss Communications Tchnology Shanghai Jiao Tong Univrsity China Lctur 5- Nov. 05 0 Classical Information Thory Sourc

More information

Supplementary Materials

Supplementary Materials 6 Supplmntary Matrials APPENDIX A PHYSICAL INTERPRETATION OF FUEL-RATE-SPEED FUNCTION A truck running on a road with grad/slop θ positiv if moving up and ngativ if moving down facs thr rsistancs: arodynamic

More information

CPSC 665 : An Algorithmist s Toolkit Lecture 4 : 21 Jan Linear Programming

CPSC 665 : An Algorithmist s Toolkit Lecture 4 : 21 Jan Linear Programming CPSC 665 : An Algorithmist s Toolkit Lctur 4 : 21 Jan 2015 Lcturr: Sushant Sachdva Linar Programming Scrib: Rasmus Kyng 1. Introduction An optimization problm rquirs us to find th minimum or maximum) of

More information

LINEAR DELAY DIFFERENTIAL EQUATION WITH A POSITIVE AND A NEGATIVE TERM

LINEAR DELAY DIFFERENTIAL EQUATION WITH A POSITIVE AND A NEGATIVE TERM Elctronic Journal of Diffrntial Equations, Vol. 2003(2003), No. 92, pp. 1 6. ISSN: 1072-6691. URL: http://jd.math.swt.du or http://jd.math.unt.du ftp jd.math.swt.du (login: ftp) LINEAR DELAY DIFFERENTIAL

More information

cycle that does not cross any edges (including its own), then it has at least

cycle that does not cross any edges (including its own), then it has at least W prov th following thorm: Thorm If a K n is drawn in th plan in such a way that it has a hamiltonian cycl that dos not cross any dgs (including its own, thn it has at last n ( 4 48 π + O(n crossings Th

More information

1 Minimum Cut Problem

1 Minimum Cut Problem CS 6 Lctur 6 Min Cut and argr s Algorithm Scribs: Png Hui How (05), Virginia Dat: May 4, 06 Minimum Cut Problm Today, w introduc th minimum cut problm. This problm has many motivations, on of which coms

More information

Introduction to Arithmetic Geometry Fall 2013 Lecture #20 11/14/2013

Introduction to Arithmetic Geometry Fall 2013 Lecture #20 11/14/2013 18.782 Introduction to Arithmtic Gomtry Fall 2013 Lctur #20 11/14/2013 20.1 Dgr thorm for morphisms of curvs Lt us rstat th thorm givn at th nd of th last lctur, which w will now prov. Thorm 20.1. Lt φ:

More information

On the irreducibility of some polynomials in two variables

On the irreducibility of some polynomials in two variables ACTA ARITHMETICA LXXXII.3 (1997) On th irrducibility of som polynomials in two variabls by B. Brindza and Á. Pintér (Dbrcn) To th mmory of Paul Erdős Lt f(x) and g(y ) b polynomials with intgral cofficints

More information

On spanning trees and cycles of multicolored point sets with few intersections

On spanning trees and cycles of multicolored point sets with few intersections On spanning trs and cycls of multicolord point sts with fw intrsctions M. Kano, C. Mrino, and J. Urrutia April, 00 Abstract Lt P 1,..., P k b a collction of disjoint point sts in R in gnral position. W

More information

COUNTING TAMELY RAMIFIED EXTENSIONS OF LOCAL FIELDS UP TO ISOMORPHISM

COUNTING TAMELY RAMIFIED EXTENSIONS OF LOCAL FIELDS UP TO ISOMORPHISM COUNTING TAMELY RAMIFIED EXTENSIONS OF LOCAL FIELDS UP TO ISOMORPHISM Jim Brown Dpartmnt of Mathmatical Scincs, Clmson Univrsity, Clmson, SC 9634, USA jimlb@g.clmson.du Robrt Cass Dpartmnt of Mathmatics,

More information

Addition of angular momentum

Addition of angular momentum Addition of angular momntum April, 07 Oftn w nd to combin diffrnt sourcs of angular momntum to charactriz th total angular momntum of a systm, or to divid th total angular momntum into parts to valuat

More information

The Equitable Dominating Graph

The Equitable Dominating Graph Intrnational Journal of Enginring Rsarch and Tchnology. ISSN 0974-3154 Volum 8, Numbr 1 (015), pp. 35-4 Intrnational Rsarch Publication Hous http://www.irphous.com Th Equitabl Dominating Graph P.N. Vinay

More information

Limiting value of higher Mahler measure

Limiting value of higher Mahler measure Limiting valu of highr Mahlr masur Arunabha Biswas a, Chris Monico a, a Dpartmnt of Mathmatics & Statistics, Txas Tch Univrsity, Lubbock, TX 7949, USA Abstract W considr th k-highr Mahlr masur m k P )

More information

(Upside-Down o Direct Rotation) β - Numbers

(Upside-Down o Direct Rotation) β - Numbers Amrican Journal of Mathmatics and Statistics 014, 4(): 58-64 DOI: 10593/jajms0140400 (Upsid-Down o Dirct Rotation) β - Numbrs Ammar Sddiq Mahmood 1, Shukriyah Sabir Ali,* 1 Dpartmnt of Mathmatics, Collg

More information

Abstract Interpretation: concrete and abstract semantics

Abstract Interpretation: concrete and abstract semantics Abstract Intrprtation: concrt and abstract smantics Concrt smantics W considr a vry tiny languag that manags arithmtic oprations on intgrs valus. Th (concrt) smantics of th languags cab b dfind by th funzcion

More information

Finding low cost TSP and 2-matching solutions using certain half integer subtour vertices

Finding low cost TSP and 2-matching solutions using certain half integer subtour vertices Finding low cost TSP and 2-matching solutions using crtain half intgr subtour vrtics Sylvia Boyd and Robrt Carr Novmbr 996 Introduction Givn th complt graph K n = (V, E) on n nods with dg costs c R E,

More information

Construction of asymmetric orthogonal arrays of strength three via a replacement method

Construction of asymmetric orthogonal arrays of strength three via a replacement method isid/ms/26/2 Fbruary, 26 http://www.isid.ac.in/ statmath/indx.php?modul=prprint Construction of asymmtric orthogonal arrays of strngth thr via a rplacmnt mthod Tian-fang Zhang, Qiaoling Dng and Alok Dy

More information

MATH 319, WEEK 15: The Fundamental Matrix, Non-Homogeneous Systems of Differential Equations

MATH 319, WEEK 15: The Fundamental Matrix, Non-Homogeneous Systems of Differential Equations MATH 39, WEEK 5: Th Fundamntal Matrix, Non-Homognous Systms of Diffrntial Equations Fundamntal Matrics Considr th problm of dtrmining th particular solution for an nsmbl of initial conditions For instanc,

More information

Exam 1. It is important that you clearly show your work and mark the final answer clearly, closed book, closed notes, no calculator.

Exam 1. It is important that you clearly show your work and mark the final answer clearly, closed book, closed notes, no calculator. Exam N a m : _ S O L U T I O N P U I D : I n s t r u c t i o n s : It is important that you clarly show your work and mark th final answr clarly, closd book, closd nots, no calculator. T i m : h o u r

More information

Brief Introduction to Statistical Mechanics

Brief Introduction to Statistical Mechanics Brif Introduction to Statistical Mchanics. Purpos: Ths nots ar intndd to provid a vry quick introduction to Statistical Mchanics. Th fild is of cours far mor vast than could b containd in ths fw pags.

More information

Application of Vague Soft Sets in students evaluation

Application of Vague Soft Sets in students evaluation Availabl onlin at www.plagiarsarchlibrary.com Advancs in Applid Scinc Rsarch, 0, (6):48-43 ISSN: 0976-860 CODEN (USA): AASRFC Application of Vagu Soft Sts in studnts valuation B. Chtia*and P. K. Das Dpartmnt

More information

Addition of angular momentum

Addition of angular momentum Addition of angular momntum April, 0 Oftn w nd to combin diffrnt sourcs of angular momntum to charactriz th total angular momntum of a systm, or to divid th total angular momntum into parts to valuat th

More information

From Elimination to Belief Propagation

From Elimination to Belief Propagation School of omputr Scinc Th lif Propagation (Sum-Product lgorithm Probabilistic Graphical Modls (10-708 Lctur 5, Sp 31, 2007 Rcptor Kinas Rcptor Kinas Kinas X 5 ric Xing Gn G T X 6 X 7 Gn H X 8 Rading: J-hap

More information

Thus, because if either [G : H] or [H : K] is infinite, then [G : K] is infinite, then [G : K] = [G : H][H : K] for all infinite cases.

Thus, because if either [G : H] or [H : K] is infinite, then [G : K] is infinite, then [G : K] = [G : H][H : K] for all infinite cases. Homwork 5 M 373K Solutions Mark Lindbrg and Travis Schdlr 1. Prov that th ring Z/mZ (for m 0) is a fild if and only if m is prim. ( ) Proof by Contrapositiv: Hr, thr ar thr cass for m not prim. m 0: Whn

More information

Higher order derivatives

Higher order derivatives Robrto s Nots on Diffrntial Calculus Chaptr 4: Basic diffrntiation ruls Sction 7 Highr ordr drivativs What you nd to know alrady: Basic diffrntiation ruls. What you can larn hr: How to rpat th procss of

More information

That is, we start with a general matrix: And end with a simpler matrix:

That is, we start with a general matrix: And end with a simpler matrix: DIAGON ALIZATION OF THE STR ESS TEN SOR INTRO DUCTIO N By th us of Cauchy s thorm w ar abl to rduc th numbr of strss componnts in th strss tnsor to only nin valus. An additional simplification of th strss

More information

Search sequence databases 3 10/25/2016

Search sequence databases 3 10/25/2016 Sarch squnc databass 3 10/25/2016 Etrm valu distribution Ø Suppos X is a random variabl with probability dnsity function p(, w sampl a larg numbr S of indpndnt valus of X from this distribution for an

More information

CS 361 Meeting 12 10/3/18

CS 361 Meeting 12 10/3/18 CS 36 Mting 2 /3/8 Announcmnts. Homwork 4 is du Friday. If Friday is Mountain Day, homwork should b turnd in at my offic or th dpartmnt offic bfor 4. 2. Homwork 5 will b availabl ovr th wknd. 3. Our midtrm

More information

Equidistribution and Weyl s criterion

Equidistribution and Weyl s criterion Euidistribution and Wyl s critrion by Brad Hannigan-Daly W introduc th ida of a sunc of numbrs bing uidistributd (mod ), and w stat and prov a thorm of Hrmann Wyl which charactrizs such suncs. W also discuss

More information

Strongly Connected Components

Strongly Connected Components Strongly Connctd Componnts Lt G = (V, E) b a dirctd graph Writ if thr is a path from to in G Writ if and is an quivalnc rlation: implis and implis s quivalnc classs ar calld th strongly connctd componnts

More information

Problem Set 6 Solutions

Problem Set 6 Solutions 6.04/18.06J Mathmatics for Computr Scinc March 15, 005 Srini Dvadas and Eric Lhman Problm St 6 Solutions Du: Monday, March 8 at 9 PM in Room 3-044 Problm 1. Sammy th Shark is a financial srvic providr

More information

EEO 401 Digital Signal Processing Prof. Mark Fowler

EEO 401 Digital Signal Processing Prof. Mark Fowler EEO 401 Digital Signal Procssing Prof. Mark Fowlr Dtails of th ot St #19 Rading Assignmnt: Sct. 7.1.2, 7.1.3, & 7.2 of Proakis & Manolakis Dfinition of th So Givn signal data points x[n] for n = 0,, -1

More information

BINOMIAL COEFFICIENTS INVOLVING INFINITE POWERS OF PRIMES. 1. Statement of results

BINOMIAL COEFFICIENTS INVOLVING INFINITE POWERS OF PRIMES. 1. Statement of results BINOMIAL COEFFICIENTS INVOLVING INFINITE POWERS OF PRIMES DONALD M. DAVIS Abstract. If p is a prim and n a positiv intgr, lt ν p (n dnot th xponnt of p in n, and u p (n n/p νp(n th unit part of n. If α

More information

Recall that by Theorems 10.3 and 10.4 together provide us the estimate o(n2 ), S(q) q 9, q=1

Recall that by Theorems 10.3 and 10.4 together provide us the estimate o(n2 ), S(q) q 9, q=1 Chaptr 11 Th singular sris Rcall that by Thorms 10 and 104 togthr provid us th stimat 9 4 n 2 111 Rn = SnΓ 2 + on2, whr th singular sris Sn was dfind in Chaptr 10 as Sn = q=1 Sq q 9, with Sq = 1 a q gcda,q=1

More information

SECTION where P (cos θ, sin θ) and Q(cos θ, sin θ) are polynomials in cos θ and sin θ, provided Q is never equal to zero.

SECTION where P (cos θ, sin θ) and Q(cos θ, sin θ) are polynomials in cos θ and sin θ, provided Q is never equal to zero. SETION 6. 57 6. Evaluation of Dfinit Intgrals Exampl 6.6 W hav usd dfinit intgrals to valuat contour intgrals. It may com as a surpris to larn that contour intgrals and rsidus can b usd to valuat crtain

More information

Quasi-Classical States of the Simple Harmonic Oscillator

Quasi-Classical States of the Simple Harmonic Oscillator Quasi-Classical Stats of th Simpl Harmonic Oscillator (Draft Vrsion) Introduction: Why Look for Eignstats of th Annihilation Oprator? Excpt for th ground stat, th corrspondnc btwn th quantum nrgy ignstats

More information

Network Congestion Games

Network Congestion Games Ntwork Congstion Gams Assistant Profssor Tas A&M Univrsity Collg Station, TX TX Dallas Collg Station Austin Houston Bst rout dpnds on othrs Ntwork Congstion Gams Travl tim incrass with congstion Highway

More information

4.5 Minimum Spanning Tree. Chapter 4. Greedy Algorithms. Minimum Spanning Tree. Motivating application

4.5 Minimum Spanning Tree. Chapter 4. Greedy Algorithms. Minimum Spanning Tree. Motivating application 1 Chaptr. Minimum panning Tr lids by Kvin Wayn. Copyright 200 Parson-Addison Wsly. All rights rsrvd. *Adjustd by Gang Tan for C33: Algorithms at Boston Collg, Fall 0 Motivating application Minimum panning

More information

Square of Hamilton cycle in a random graph

Square of Hamilton cycle in a random graph Squar of Hamilton cycl in a random graph Andrzj Dudk Alan Friz Jun 28, 2016 Abstract W show that p = n is a sharp thrshold for th random graph G n,p to contain th squar of a Hamilton cycl. This improvs

More information

General Notes About 2007 AP Physics Scoring Guidelines

General Notes About 2007 AP Physics Scoring Guidelines AP PHYSICS C: ELECTRICITY AND MAGNETISM 2007 SCORING GUIDELINES Gnral Nots About 2007 AP Physics Scoring Guidlins 1. Th solutions contain th most common mthod of solving th fr-rspons qustions and th allocation

More information

Function Spaces. a x 3. (Letting x = 1 =)) a(0) + b + c (1) = 0. Row reducing the matrix. b 1. e 4 3. e 9. >: (x = 1 =)) a(0) + b + c (1) = 0

Function Spaces. a x 3. (Letting x = 1 =)) a(0) + b + c (1) = 0. Row reducing the matrix. b 1. e 4 3. e 9. >: (x = 1 =)) a(0) + b + c (1) = 0 unction Spacs Prrquisit: Sction 4.7, Coordinatization n this sction, w apply th tchniqus of Chaptr 4 to vctor spacs whos lmnts ar functions. Th vctor spacs P n and P ar familiar xampls of such spacs. Othr

More information

Final Exam Solutions

Final Exam Solutions CS 2 Advancd Data Structurs and Algorithms Final Exam Solutions Jonathan Turnr /8/20. (0 points) Suppos that r is a root of som tr in a Fionacci hap. Assum that just for a dltmin opration, r has no childrn

More information

COMPUTER GENERATED HOLOGRAMS Optical Sciences 627 W.J. Dallas (Monday, April 04, 2005, 8:35 AM) PART I: CHAPTER TWO COMB MATH.

COMPUTER GENERATED HOLOGRAMS Optical Sciences 627 W.J. Dallas (Monday, April 04, 2005, 8:35 AM) PART I: CHAPTER TWO COMB MATH. C:\Dallas\0_Courss\03A_OpSci_67\0 Cgh_Book\0_athmaticalPrliminaris\0_0 Combath.doc of 8 COPUTER GENERATED HOLOGRAS Optical Scincs 67 W.J. Dallas (onday, April 04, 005, 8:35 A) PART I: CHAPTER TWO COB ATH

More information

Approximate Maximum Flow in Undirected Networks by Christiano, Kelner, Madry, Spielmann, Teng (STOC 2011)

Approximate Maximum Flow in Undirected Networks by Christiano, Kelner, Madry, Spielmann, Teng (STOC 2011) Approximat Maximum Flow in Undirctd Ntworks by Christiano, Klnr, Madry, Spilmann, Tng (STOC 2011) Kurt Mhlhorn Max Planck Institut for Informatics and Saarland Univrsity Sptmbr 28, 2011 Th Rsult High-Lvl

More information

BINOMIAL COEFFICIENTS INVOLVING INFINITE POWERS OF PRIMES

BINOMIAL COEFFICIENTS INVOLVING INFINITE POWERS OF PRIMES BINOMIAL COEFFICIENTS INVOLVING INFINITE POWERS OF PRIMES DONALD M. DAVIS Abstract. If p is a prim (implicit in notation and n a positiv intgr, lt ν(n dnot th xponnt of p in n, and U(n n/p ν(n, th unit

More information

1973 AP Calculus AB: Section I

1973 AP Calculus AB: Section I 97 AP Calculus AB: Sction I 9 Minuts No Calculator Not: In this amination, ln dnots th natural logarithm of (that is, logarithm to th bas ).. ( ) d= + C 6 + C + C + C + C. If f ( ) = + + + and ( ), g=

More information

Abstract Interpretation. Lecture 5. Profs. Aiken, Barrett & Dill CS 357 Lecture 5 1

Abstract Interpretation. Lecture 5. Profs. Aiken, Barrett & Dill CS 357 Lecture 5 1 Abstract Intrprtation 1 History On brakthrough papr Cousot & Cousot 77 (?) Inspird by Dataflow analysis Dnotational smantics Enthusiastically mbracd by th community At last th functional community... At

More information

Mutually Independent Hamiltonian Cycles of Pancake Networks

Mutually Independent Hamiltonian Cycles of Pancake Networks Mutually Indpndnt Hamiltonian Cycls of Pancak Ntworks Chng-Kuan Lin Dpartmnt of Mathmatics National Cntral Univrsity, Chung-Li, Taiwan 00, R O C discipl@ms0urlcomtw Hua-Min Huang Dpartmnt of Mathmatics

More information

Homotopy perturbation technique

Homotopy perturbation technique Comput. Mthods Appl. Mch. Engrg. 178 (1999) 257±262 www.lsvir.com/locat/cma Homotopy prturbation tchniqu Ji-Huan H 1 Shanghai Univrsity, Shanghai Institut of Applid Mathmatics and Mchanics, Shanghai 272,

More information

Searching Linked Lists. Perfect Skip List. Building a Skip List. Skip List Analysis (1) Assume the list is sorted, but is stored in a linked list.

Searching Linked Lists. Perfect Skip List. Building a Skip List. Skip List Analysis (1) Assume the list is sorted, but is stored in a linked list. 3 3 4 8 6 3 3 4 8 6 3 3 4 8 6 () (d) 3 Sarching Linkd Lists Sarching Linkd Lists Sarching Linkd Lists ssum th list is sortd, but is stord in a linkd list. an w us binary sarch? omparisons? Work? What if

More information

Mor Tutorial at www.dumblittldoctor.com Work th problms without a calculator, but us a calculator to chck rsults. And try diffrntiating your answrs in part III as a usful chck. I. Applications of Intgration

More information

2.3 Matrix Formulation

2.3 Matrix Formulation 23 Matrix Formulation 43 A mor complicatd xampl ariss for a nonlinar systm of diffrntial quations Considr th following xampl Exampl 23 x y + x( x 2 y 2 y x + y( x 2 y 2 (233 Transforming to polar coordinats,

More information

ECE602 Exam 1 April 5, You must show ALL of your work for full credit.

ECE602 Exam 1 April 5, You must show ALL of your work for full credit. ECE62 Exam April 5, 27 Nam: Solution Scor: / This xam is closd-book. You must show ALL of your work for full crdit. Plas rad th qustions carfully. Plas chck your answrs carfully. Calculators may NOT b

More information

Chapter 10. The singular integral Introducing S(n) and J(n)

Chapter 10. The singular integral Introducing S(n) and J(n) Chaptr Th singular intgral Our aim in this chaptr is to rplac th functions S (n) and J (n) by mor convnint xprssions; ths will b calld th singular sris S(n) and th singular intgral J(n). This will b don

More information

Middle East Technical University Department of Mechanical Engineering ME 413 Introduction to Finite Element Analysis

Middle East Technical University Department of Mechanical Engineering ME 413 Introduction to Finite Element Analysis Middl East Tchnical Univrsity Dpartmnt of Mchanical Enginring ME 43 Introduction to Finit Elmnt Analysis Chaptr 3 Computr Implmntation of D FEM Ths nots ar prpard by Dr. Cünyt Srt http://www.m.mtu.du.tr/popl/cunyt

More information

Differential Equations

Differential Equations Prfac Hr ar m onlin nots for m diffrntial quations cours that I tach hr at Lamar Univrsit. Dspit th fact that ths ar m class nots, th should b accssibl to anon wanting to larn how to solv diffrntial quations

More information

u 3 = u 3 (x 1, x 2, x 3 )

u 3 = u 3 (x 1, x 2, x 3 ) Lctur 23: Curvilinar Coordinats (RHB 8.0 It is oftn convnint to work with variabls othr than th Cartsian coordinats x i ( = x, y, z. For xampl in Lctur 5 w mt sphrical polar and cylindrical polar coordinats.

More information

Cramér-Rao Inequality: Let f(x; θ) be a probability density function with continuous parameter

Cramér-Rao Inequality: Let f(x; θ) be a probability density function with continuous parameter WHEN THE CRAMÉR-RAO INEQUALITY PROVIDES NO INFORMATION STEVEN J. MILLER Abstract. W invstigat a on-paramtr family of probability dnsitis (rlatd to th Parto distribution, which dscribs many natural phnomna)

More information

Electrical Flows, Laplacian Systems, and Faster Approximation of Maximum Flow in Undirected Graphs

Electrical Flows, Laplacian Systems, and Faster Approximation of Maximum Flow in Undirected Graphs Elctrical Flows, Laplacian Systms, and Fastr Approximation of Maximum Flow in Undirctd Graphs Paul Christiano Mathmatics MIT paulfchristiano@gmail.com Danil A. Spilman Computr Scinc Yal Univrsity spilman@cs.yal.du

More information

Propositional Logic. Combinatorial Problem Solving (CPS) Albert Oliveras Enric Rodríguez-Carbonell. May 17, 2018

Propositional Logic. Combinatorial Problem Solving (CPS) Albert Oliveras Enric Rodríguez-Carbonell. May 17, 2018 Propositional Logic Combinatorial Problm Solving (CPS) Albrt Olivras Enric Rodríguz-Carbonll May 17, 2018 Ovrviw of th sssion Dfinition of Propositional Logic Gnral Concpts in Logic Rduction to SAT CNFs

More information

Section 11.6: Directional Derivatives and the Gradient Vector

Section 11.6: Directional Derivatives and the Gradient Vector Sction.6: Dirctional Drivativs and th Gradint Vctor Practic HW rom Stwart Ttbook not to hand in p. 778 # -4 p. 799 # 4-5 7 9 9 35 37 odd Th Dirctional Drivativ Rcall that a b Slop o th tangnt lin to th

More information

Einstein Equations for Tetrad Fields

Einstein Equations for Tetrad Fields Apiron, Vol 13, No, Octobr 006 6 Einstin Equations for Ttrad Filds Ali Rıza ŞAHİN, R T L Istanbul (Turky) Evry mtric tnsor can b xprssd by th innr product of ttrad filds W prov that Einstin quations for

More information

Hardy-Littlewood Conjecture and Exceptional real Zero. JinHua Fei. ChangLing Company of Electronic Technology Baoji Shannxi P.R.

Hardy-Littlewood Conjecture and Exceptional real Zero. JinHua Fei. ChangLing Company of Electronic Technology Baoji Shannxi P.R. Hardy-Littlwood Conjctur and Excptional ral Zro JinHua Fi ChangLing Company of Elctronic Tchnology Baoji Shannxi P.R.China E-mail: fijinhuayoujian@msn.com Abstract. In this papr, w assum that Hardy-Littlwood

More information

Section 6.1. Question: 2. Let H be a subgroup of a group G. Then H operates on G by left multiplication. Describe the orbits for this operation.

Section 6.1. Question: 2. Let H be a subgroup of a group G. Then H operates on G by left multiplication. Describe the orbits for this operation. MAT 444 H Barclo Spring 004 Homwork 6 Solutions Sction 6 Lt H b a subgroup of a group G Thn H oprats on G by lft multiplication Dscrib th orbits for this opration Th orbits of G ar th right costs of H

More information

INCOMPLETE KLOOSTERMAN SUMS AND MULTIPLICATIVE INVERSES IN SHORT INTERVALS. xy 1 (mod p), (x, y) I (j)

INCOMPLETE KLOOSTERMAN SUMS AND MULTIPLICATIVE INVERSES IN SHORT INTERVALS. xy 1 (mod p), (x, y) I (j) INCOMPLETE KLOOSTERMAN SUMS AND MULTIPLICATIVE INVERSES IN SHORT INTERVALS T D BROWNING AND A HAYNES Abstract W invstigat th solubility of th congrunc xy (mod ), whr is a rim and x, y ar rstrictd to li

More information

Two Products Manufacturer s Production Decisions with Carbon Constraint

Two Products Manufacturer s Production Decisions with Carbon Constraint Managmnt Scinc and Enginring Vol 7 No 3 pp 3-34 DOI:3968/jms9335X374 ISSN 93-34 [Print] ISSN 93-35X [Onlin] wwwcscanadant wwwcscanadaorg Two Products Manufacturr s Production Dcisions with Carbon Constraint

More information

Deift/Zhou Steepest descent, Part I

Deift/Zhou Steepest descent, Part I Lctur 9 Dift/Zhou Stpst dscnt, Part I W now focus on th cas of orthogonal polynomials for th wight w(x) = NV (x), V (x) = t x2 2 + x4 4. Sinc th wight dpnds on th paramtr N N w will writ π n,n, a n,n,

More information

On the number of pairs of positive integers x,y H such that x 2 +y 2 +1, x 2 +y 2 +2 are square-free

On the number of pairs of positive integers x,y H such that x 2 +y 2 +1, x 2 +y 2 +2 are square-free arxiv:90.04838v [math.nt] 5 Jan 09 On th numbr of pairs of positiv intgrs x,y H such that x +y +, x +y + ar squar-fr S. I. Dimitrov Abstract In th prsnt papr w show that thr xist infinitly many conscutiv

More information

Approximation and Inapproximation for The Influence Maximization Problem in Social Networks under Deterministic Linear Threshold Model

Approximation and Inapproximation for The Influence Maximization Problem in Social Networks under Deterministic Linear Threshold Model 20 3st Intrnational Confrnc on Distributd Computing Systms Workshops Approximation and Inapproximation for Th Influnc Maximization Problm in Social Ntworks undr Dtrministic Linar Thrshold Modl Zaixin Lu,

More information

2008 AP Calculus BC Multiple Choice Exam

2008 AP Calculus BC Multiple Choice Exam 008 AP Multipl Choic Eam Nam 008 AP Calculus BC Multipl Choic Eam Sction No Calculator Activ AP Calculus 008 BC Multipl Choic. At tim t 0, a particl moving in th -plan is th acclration vctor of th particl

More information

International Journal of Scientific & Engineering Research, Volume 6, Issue 7, July ISSN

International Journal of Scientific & Engineering Research, Volume 6, Issue 7, July ISSN Intrnational Journal of Scintific & Enginring Rsarch, Volum 6, Issu 7, July-25 64 ISSN 2229-558 HARATERISTIS OF EDGE UTSET MATRIX OF PETERSON GRAPH WITH ALGEBRAI GRAPH THEORY Dr. G. Nirmala M. Murugan

More information

Roadmap. XML Indexing. DataGuide example. DataGuides. Strong DataGuides. Multiple DataGuides for same data. CPS Topics in Database Systems

Roadmap. XML Indexing. DataGuide example. DataGuides. Strong DataGuides. Multiple DataGuides for same data. CPS Topics in Database Systems Roadmap XML Indxing CPS 296.1 Topics in Databas Systms Indx fabric Coopr t al. A Fast Indx for Smistructurd Data. VLDB, 2001 DataGuid Goldman and Widom. DataGuids: Enabling Qury Formulation and Optimization

More information

u x v x dx u x v x v x u x dx d u x v x u x v x dx u x v x dx Integration by Parts Formula

u x v x dx u x v x v x u x dx d u x v x u x v x dx u x v x dx Integration by Parts Formula 7. Intgration by Parts Each drivativ formula givs ris to a corrsponding intgral formula, as w v sn many tims. Th drivativ product rul yilds a vry usful intgration tchniqu calld intgration by parts. Starting

More information

Vishnu V. Narayan. January

Vishnu V. Narayan. January A 17 12 -approimation algorithm for 2-rt-connctd spanning subgraphs on graphs with minimum dgr at last arxi:1612.047902 [cs.ds] 17 Jan 2017 Vishnu V. Naraan Januar 17 2017 W obtain a polnomial-tim 17 -approimation

More information

Electrical Flows, Laplacian Systems, and Faster Approximation of Maximum Flow in Undirected Graphs

Electrical Flows, Laplacian Systems, and Faster Approximation of Maximum Flow in Undirected Graphs Elctrical Flows, Laplacian Systms, and Fastr Approximation of Maximum Flow in Undirctd Graphs Paul Christiano MIT Jonathan A. Klnr MIT Alksandr Mądry MIT Shang-Hua Tng Univrsity of Southrn California Octobr

More information

ME 321 Kinematics and Dynamics of Machines S. Lambert Winter 2002

ME 321 Kinematics and Dynamics of Machines S. Lambert Winter 2002 3.4 Forc Analysis of Linkas An undrstandin of forc analysis of linkas is rquird to: Dtrmin th raction forcs on pins, tc. as a consqunc of a spcifid motion (don t undrstimat th sinificanc of dynamic or

More information

A Propagating Wave Packet Group Velocity Dispersion

A Propagating Wave Packet Group Velocity Dispersion Lctur 8 Phys 375 A Propagating Wav Packt Group Vlocity Disprsion Ovrviw and Motivation: In th last lctur w lookd at a localizd solution t) to th 1D fr-particl Schrödingr quation (SE) that corrsponds to

More information

EXST Regression Techniques Page 1

EXST Regression Techniques Page 1 EXST704 - Rgrssion Tchniqus Pag 1 Masurmnt rrors in X W hav assumd that all variation is in Y. Masurmnt rror in this variabl will not ffct th rsults, as long as thy ar uncorrlatd and unbiasd, sinc thy

More information

A. Limits and Horizontal Asymptotes ( ) f x f x. f x. x "±# ( ).

A. Limits and Horizontal Asymptotes ( ) f x f x. f x. x ±# ( ). A. Limits and Horizontal Asymptots What you ar finding: You can b askd to find lim x "a H.A.) problm is asking you find lim x "# and lim x "$#. or lim x "±#. Typically, a horizontal asymptot algbraically,

More information

1 N N(θ;d 1...d l ;N) 1 q l = o(1)

1 N N(θ;d 1...d l ;N) 1 q l = o(1) NORMALITY OF NUMBERS GENERATED BY THE VALUES OF ENTIRE FUNCTIONS MANFRED G. MADRITSCH, JÖRG M. THUSWALDNER, AND ROBERT F. TICHY Abstract. W show that th numbr gnratd by th q-ary intgr part of an ntir function

More information

Some remarks on Kurepa s left factorial

Some remarks on Kurepa s left factorial Som rmarks on Kurpa s lft factorial arxiv:math/0410477v1 [math.nt] 21 Oct 2004 Brnd C. Kllnr Abstract W stablish a connction btwn th subfactorial function S(n) and th lft factorial function of Kurpa K(n).

More information

4.5 Minimum Spanning Tree. Chapter 4. Greedy Algorithms. Minimum Spanning Tree. Applications

4.5 Minimum Spanning Tree. Chapter 4. Greedy Algorithms. Minimum Spanning Tree. Applications Chaptr. Minimum panning Tr Grdy Algorithms lids by Kvin Wayn. Copyright 200 Parson-Addison Wsly. All rights rsrvd. Minimum panning Tr Applications Minimum spanning tr. Givn a connctd graph G = (V, E) with

More information

Another view for a posteriori error estimates for variational inequalities of the second kind

Another view for a posteriori error estimates for variational inequalities of the second kind Accptd by Applid Numrical Mathmatics in July 2013 Anothr viw for a postriori rror stimats for variational inqualitis of th scond kind Fi Wang 1 and Wimin Han 2 Abstract. In this papr, w giv anothr viw

More information

Combinatorial Networks Week 1, March 11-12

Combinatorial Networks Week 1, March 11-12 1 Nots on March 11 Combinatorial Ntwors W 1, March 11-1 11 Th Pigonhol Principl Th Pigonhol Principl If n objcts ar placd in hols, whr n >, thr xists a box with mor than on objcts 11 Thorm Givn a simpl

More information

Symmetric centrosymmetric matrix vector multiplication

Symmetric centrosymmetric matrix vector multiplication Linar Algbra and its Applications 320 (2000) 193 198 www.lsvir.com/locat/laa Symmtric cntrosymmtric matrix vctor multiplication A. Mlman 1 Dpartmnt of Mathmatics, Univrsity of San Francisco, San Francisco,

More information

Hydrogen Atom and One Electron Ions

Hydrogen Atom and One Electron Ions Hydrogn Atom and On Elctron Ions Th Schrödingr quation for this two-body problm starts out th sam as th gnral two-body Schrödingr quation. First w sparat out th motion of th cntr of mass. Th intrnal potntial

More information

UNTYPED LAMBDA CALCULUS (II)

UNTYPED LAMBDA CALCULUS (II) 1 UNTYPED LAMBDA CALCULUS (II) RECALL: CALL-BY-VALUE O.S. Basic rul Sarch ruls: (\x.) v [v/x] 1 1 1 1 v v CALL-BY-VALUE EVALUATION EXAMPLE (\x. x x) (\y. y) x x [\y. y / x] = (\y. y) (\y. y) y [\y. y /

More information

Fourier Transforms and the Wave Equation. Key Mathematics: More Fourier transform theory, especially as applied to solving the wave equation.

Fourier Transforms and the Wave Equation. Key Mathematics: More Fourier transform theory, especially as applied to solving the wave equation. Lur 7 Fourir Transforms and th Wav Euation Ovrviw and Motivation: W first discuss a fw faturs of th Fourir transform (FT), and thn w solv th initial-valu problm for th wav uation using th Fourir transform

More information

5.80 Small-Molecule Spectroscopy and Dynamics

5.80 Small-Molecule Spectroscopy and Dynamics MIT OpnCoursWar http://ocw.mit.du 5.80 Small-Molcul Spctroscopy and Dynamics Fall 008 For information about citing ths matrials or our Trms of Us, visit: http://ocw.mit.du/trms. Lctur # 3 Supplmnt Contnts

More information

Solution of Assignment #2

Solution of Assignment #2 olution of Assignmnt #2 Instructor: Alirza imchi Qustion #: For simplicity, assum that th distribution function of T is continuous. Th distribution function of R is: F R ( r = P( R r = P( log ( T r = P(log

More information

Stochastic Submodular Maximization

Stochastic Submodular Maximization Stochastic Submodular Maximization Arash Asadpour, Hamid Nazrzadh, and Amin Sabri Stanford Univrsity, Stanford, CA. {asadpour,hamidnz,sabri}@stanford.du Abstract. W study stochastic submodular maximization

More information

Estimation of apparent fraction defective: A mathematical approach

Estimation of apparent fraction defective: A mathematical approach Availabl onlin at www.plagiarsarchlibrary.com Plagia Rsarch Library Advancs in Applid Scinc Rsarch, 011, (): 84-89 ISSN: 0976-8610 CODEN (USA): AASRFC Estimation of apparnt fraction dfctiv: A mathmatical

More information

Mathematics. Complex Number rectangular form. Quadratic equation. Quadratic equation. Complex number Functions: sinusoids. Differentiation Integration

Mathematics. Complex Number rectangular form. Quadratic equation. Quadratic equation. Complex number Functions: sinusoids. Differentiation Integration Mathmatics Compl numbr Functions: sinusoids Sin function, cosin function Diffrntiation Intgration Quadratic quation Quadratic quations: a b c 0 Solution: b b 4ac a Eampl: 1 0 a= b=- c=1 4 1 1or 1 1 Quadratic

More information

Properties of Phase Space Wavefunctions and Eigenvalue Equation of Momentum Dispersion Operator

Properties of Phase Space Wavefunctions and Eigenvalue Equation of Momentum Dispersion Operator Proprtis of Phas Spac Wavfunctions and Eignvalu Equation of Momntum Disprsion Oprator Ravo Tokiniaina Ranaivoson 1, Raolina Andriambololona 2, Hanitriarivo Rakotoson 3 raolinasp@yahoo.fr 1 ;jacqulinraolina@hotmail.com

More information