Compressed Suffix Trees in Practice

Size: px
Start display at page:

Download "Compressed Suffix Trees in Practice"

Transcription

1 Introdction CST design CST in practice Copressed Sffix Trees in Practice Sion Gog Copting and Inforation Systes The University of Melborne Febrary 13th 213

2 Introdction CST design CST in practice Otline 1 Introdction Basic data strctres The sffix tree 2 CST design NAV (tree topology and navigation) CSA (lexicographic inforation) LCP (longest coon prefixes) 3 CST in practice The sdsl library

3 Introdction CST design CST in practice Sccinct data strctres (1) Data strctre D representation of an object X Exaple: Rank-bit-vector bit vector b of length n (,1,,1,1,,1,1) (,,1,1,2,3,3,4) in n bits space + + operations on X access b[i] in O(1) tie rank(i) = i 1 j= b[j] in O(n) tie Sccinct data strctre D Space of D is close the inforation theoretic lower bond to represent X, while operations can still be perfored efficient.

4 Introdction CST design CST in practice Sccinct data strctres (1) Data strctre D representation of an object X Exaple: Rank-bit-vector bit vector b of length n (,1,,1,1,,1,1) (,,1,1,2,3,3,4) in n bits space + + operations on X access b[i] in O(1) tie rank(i) = i 1 j= b[j] in O(n) tie Sccinct data strctre D Space of D is close the inforation theoretic lower bond to represent X, while operations can still be perfored efficient.

5 Introdction CST design CST in practice Sccinct data strctres (1) Data strctre D representation of an object X Exaple: Rank-bit-vector bit vector b of length n (,1,,1,1,,1,1) (,,1,1,2,3,3,4) in n + n log n bits space + + operations on X access b[i] in O(1) tie rank(i) = i 1 j= b[j] in O(1) tie Sccinct data strctre D Space of D is close the inforation theoretic lower bond to represent X, while operations can still be perfored efficient.

6 Introdction CST design CST in practice Sccinct data strctres (1) Data strctre D representation of an object X Exaple: Rank-bit-vector bit vector b of length n (,1,,1,1,,1,1) (,,1,1,2,3,3,4) in n + n log n bits space + + operations on X access b[i] in O(1) tie rank(i) = i 1 j= b[j] in O(1) tie Sccinct data strctre D Space of D is close the inforation theoretic lower bond to represent X, while operations can still be perfored efficient.

7 Introdction CST design CST in practice Sccinct data strctres (2) Can sccinct data strctres replace classic ncopressed data strctres in practice? Less eory fewer CPU cycles!? Less eory less costs!? Probles: in theory develop sccinct data strctres in practice constants in O(1)-tie ters are large o(n)-space ter is not negligible coplex data strctres are hard to ipleent

8 Introdction CST design CST in practice Sccinct data strctres (2) Can sccinct data strctres replace classic ncopressed data strctres in practice? Less eory fewer CPU cycles!? 1 CPU cycle 5 CPU cycles CPU L1-Cache L2-Cache L3-Cache DRAM Disk 1 B 1 KB 512 KB 1-8 MB 4 GB x 1 GB Less eory less costs!? Probles: in theory develop sccinct data strctres in practice constants in O(1)-tie ters are large o(n)-space ter is not negligible

9 Introdction CST design CST in practice Sccinct data strctres (2) Can sccinct data strctres replace classic ncopressed data strctres in practice? Less eory fewer CPU cycles!? Less eory less costs!? Instance nae ain eory price per hor Micro 613. MB.2 US High-Meory Qadrple Extra Large 68.4 GB 2. US Pricing of Aazons Elastic Clod Copting (EC2) service in Jly 211. Probles: in theory develop sccinct data strctres in practice constants in O(1)-tie ters are large o(n)-space ter is not negligible coplex data strctres are hard to ipleent

10 Introdction CST design CST in practice Sccinct data strctres (2) Can sccinct data strctres replace classic ncopressed data strctres in practice? Less eory fewer CPU cycles!? Less eory less costs!? Probles: in theory develop sccinct data strctres in practice constants in O(1)-tie ters are large o(n)-space ter is not negligible coplex data strctres are hard to ipleent

11 Introdction CST design CST in practice The classic index data strctre: The sffix tree (ST) Let T be a text of length n over alphabet Σ of size σ. Sffix tree index data strctre for T (constrction O(n)) can be sed to solve any probles in optial tie coplexity bioinforatics data copression ses O(n log n) bits! In practice (ASCII-alphabet) 17 ties the size of T Can not handle The Attack of Massive Data DNA seqencing data (NGS)...

12 Introdction CST design CST in practice Exaple: ST of T=lndl 15 7 dl ndl l 3 ndl l 1 ndl ndl ndl l 13 ndl l 5 n = 16 Σ ={,d,l,,n,} σ = ndl Classic ipleentation ses pointers each of size 4 or 8 bytes!

13 Introdction CST design CST in practice Exaple: ST of T=lndl 15 7 dl ndl l l ndl ndl ndl ndl l 13 ndl 8 l ndl Operations root() is leaf (v) parent(v) degree(v) 5 child(v, c) select child(v, i) depth(v) edge(v, d) lca(v, w) sl(v) wl(v, c)

14 Introdction CST design CST in practice CSTs Goal of a CST ipleentation Replace fastest ncopressed ST ipleentations in different scenarios (a) both fit in RAM and we easre tie (b) both fit in RAM and we easre resorce costs (c) only CST fits in RAM and we easre tie Proposals Sadakane s CST cst_sada Flly Copressed Sffix Tree (Rsso et al.) CSTs based on interval representation of nodes (Fischer et al. csty, Ohlebsch et al. cst_sct3)

15 Introdction CST design CST in practice CSTs Goal of a CST ipleentation Replace fastest ncopressed ST ipleentations in different scenarios (a) both fit in RAM and we easre tie (b) both fit in RAM and we easre resorce costs (c) only CST fits in RAM and we easre tie Proposals which ight work for (a) and (b) Sadakane s CST cst_sada Flly Copressed Sffix Tree (Rsso et al.) CSTs based on interval representation of nodes (Fischer et al. csty, Ohlebsch et al. cst_sct3)

16 Introdction CST design CST in practice Otline 1 Introdction Basic data strctres The sffix tree 2 CST design NAV (tree topology and navigation) CSA (lexicographic inforation) LCP (longest coon prefixes) 3 CST in practice The sdsl library

17 Introdction CST design CST in practice Big pictre of CST design Ψ Wavelet Tree T BWT Hffan excess CSA NAV Min-Max-Tree LF Pioneer RMQ arentheses Seq 2n bits first child LCP PLCP

18 Introdction CST design CST in practice Big pictre of CST design 15 7 dl 14 l ndl l ndl 11 ndl 9 l 3 ndl ndl ndl l ndl (()()(()())(()((()())()()))()((()())(()(()()))()))

19 Introdction CST design CST in practice Exaple: Copressing NAV 15 7 dl l l ndl ndl l l ndl ndl 6 9 ndl 3 12 ndl ndl 11 1 tree ncopressed O(n log n) bits BPS dfs = (()()(()())(()((()())()()))()((()())(()(()()))())) copressed 4n bits

20 Introdction CST design CST in practice Exaple: Copressing NAV 15 7 dl l l ndl ndl l l ndl ndl 6 9 ndl 3 12 ndl ndl 11 1 tree ncopressed O(n log n) bits BPS dfs = (()()(()())(()((()())()()))()((()())(()(()()))())) copressed 4n bits

21 Introdction CST design CST in practice Exaple: Copressing NAV dl l l ndl ndl l l ndl ndl 6 9 ndl 3 12 ndl ndl 11 1 tree ncopressed O(n log n) bits BPS dfs = (()()(()())(()((()())()()))()((()())(()(()()))())) copressed 4n bits

22 Introdction CST design CST in practice Exaple: Copressing NAV dl l l ndl ndl l l ndl ndl 6 9 ndl 3 12 ndl ndl 11 1 tree ncopressed O(n log n) bits BPS dfs = (()()(()())(()((()())()()))()((()())(()(()()))())) copressed 4n bits

23 Introdction CST design CST in practice Exaple: Copressing NAV dl l l ndl ndl l l ndl ndl 6 9 ndl 3 12 ndl ndl 11 1 tree ncopressed O(n log n) bits BPS dfs = (()()(()())(()((()())()()))()((()())(()(()()))())) copressed 4n bits

24 Introdction CST design CST in practice Exaple: Copressing NAV dl l l ndl ndl l l ndl ndl 6 9 ndl 3 12 ndl ndl 11 1 tree ncopressed O(n log n) bits BPS dfs = (()()(()())(()((()())()()))()((()())(()(()()))())) copressed 4n bits

25 Introdction CST design CST in practice Exaple: Copressing NAV 15 1 dl l l 5 ndl ndl ndl 11 l l ndl ndl 6 9 ndl 3 12 ndl tree ncopressed O(n log n) bits BPS dfs = (()()(()())(()((()())()()))()((()())(()(()()))())) copressed 4n bits

26 Introdction CST design CST in practice Exaple: Copressing NAV 15 1 dl l l 5 ndl ndl ndl 11 l 6 l ndl ndl 6 9 ndl 3 12 ndl tree ncopressed O(n log n) bits BPS dfs = (()()(()())(()((()())()()))()((()())(()(()()))())) copressed 4n bits

27 Introdction CST design CST in practice Exaple: Copressing NAV 15 1 dl l l 5 ndl ndl ndl 11 l 6 l ndl ndl 6 9 ndl 3 12 ndl tree ncopressed O(n log n) bits BPS dfs = (()()(()())(()((()())()()))()((()())(()(()()))())) copressed 4n bits

28 Introdction CST design CST in practice Exaple: Copressing NAV 15 1 dl l l 5 ndl ndl ndl 11 l 6 l ndl ndl 6 9 ndl ndl tree ncopressed O(n log n) bits BPS dfs = (()()(()())(()((()())()()))()((()())(()(()()))())) copressed 4n bits

29 Introdction CST design CST in practice Exaple: Copressing NAV 15 1 dl l l 5 ndl ndl ndl 11 l 6 l ndl ndl 6 9 ndl ndl tree ncopressed O(n log n) bits BPS dfs = (()()(()())(()((()())()()))()((()())(()(()()))())) copressed 4n bits

30 Introdction CST design CST in practice Exaple: Copressing NAV 15 1 dl l l 5 11 ndl ndl ndl 11 l 36 6 l 15 ndl ndl ndl ndl tree ncopressed O(n log n) bits BPS dfs = (()()(()())(()((()())()()))()((()())(()(()()))())) copressed 4n bits

31 Introdction CST design CST in practice NAV data strctres (1) 4n bits 2n bits 15 1 dl l l 5 11 ndl ndl ndl 11 l 36 6 l 15 ndl ndl ndl ndl BPS dfs = (()()(()())(()((()())()()))()((()())(()(()()))())) -[, 15] 15 dl 1-[1, 15] l 73-[2, 3] 1-[4, 14 8] l ndl ndl 2-[5, 8] 4-[1, 11] ndl 11 l 2-[12, 14] 5-[5, 6] l ndl ndl 6-[13, 14] 6 9 ndl 3 12 ndl BPS sct= ( ( ( ( )( ( ( )( ( )))( ( ( )( ( ( ))( )))))))) LCP= o(n) bits to answer find open(i), find close(i), enclose(i), doble enclose(i, j),rank(i, c), select(i, c),... in constant tie

32 Introdction CST design CST in practice NAV data strctres (2) Coparison of different NAV strctres cst_sada cst_sct cst_sct3 space in bits 4n + o(n) 2n + o(n) 3n + o(n) root() O(1) O(1) O(1) degree(v) O(σ) O(t LCP log σ) O(1) depth(v) O(t LCP ) O(t LCP ) O(t LCP ) parent(v) O(1) O(t LCP log σ) O(1) select child(v, i) O(i) O(t LCP ) O(1) sibling(v) O(1) O(t LCP ) O(1) sl(v), lca(v, w) O(1) O(t LCP log σ) O(1) child(v, c) O(t SA σ) O(t SA log σ) O(t SA log σ)

33 Introdction CST design CST in practice Exaple operations: select leaf (i) and lca(v, w) on NAV 15 7 dl l l ndl ndl l l ndl ndl 6 9 ndl ndl ndl 11 BPS dfs = (()()(()())(()((()())()()))()((()())(()(()()))())) 1 select leaf (4) = select(4, 1 ) = 8 select leaf (5) = select(5, 1 ) = 12 lca(8,12) = doble enclose(8,12) =

34 Introdction CST design CST in practice Virtes of a CSA based on BWT Sall size: CSA = BWT + n log n s SA where BWT can be chosen to be n log σ bits nh (T) bits nh k (T) + O(σ k ) bits bits pattern atching in tie O( P log σ) (even O( P ) for σ polylog(n)) by backward search (Ferragina & Manzini)

35 Introdction CST design CST in practice H k of the Pizza&Chili 2MB test cases dblp.xl dna english proteins rand_k128 sorces k H k CT /n H k CT /n H k CT /n H k CT /n H k CT /n H k CT /n

36 Introdction CST design CST in practice Backward search i T BWT n l l d F dl l lndl l lndl ndl ndl l lndl l lndl ndl (a) T BWT n l l d F dl l lndl l lndl ndl ndl l lndl l lndl ndl (b) T BWT n l l d F dl l lndl l lndl ndl ndl l lndl l lndl ndl (c)

37 Introdction CST design CST in practice Wavelet tree: rank for character seqences d 1 nlld lld 11 1 ll 1 n n 1 1 nlld nlld nlld d 1 n d 1 1 nll (a) d ll n (b)

38 Introdction CST design CST in practice Exaple: Copressing CSA (practical approach) i SA LF T BWT T n l l d dl l lndl l lndl ndl ndl l lndl l lndl ndl

39 Introdction CST design CST in practice Exaple: Copressing CSA (practical approach) SA ncopressed n log n bits n l l d copressed d l l n s SA = 3 access LF[i] in tie O(log σ) access CSA[i] in tie O(s SA log σ) CSA = T BWT + SA saples: n log σ + o(n log σ) bits + n log n s SA bits

40 Introdction CST design CST in practice Exaple: Copressing CSA (practical approach) SA[13]= n l d l l n s SA = 3 access LF[i] in tie O(log σ) access CSA[i] in tie O(s SA log σ) l d SA ncopressed n log n bits copressed CSA = T BWT + SA saples: n log σ + o(n log σ) bits + n log n s SA bits

41 Introdction CST design CST in practice Exaple: Copressing CSA (practical approach) SA[13]=SA[1]+1 n l d l l n s SA = 3 access LF[i] in tie O(log σ) access CSA[i] in tie O(s SA log σ) l d SA ncopressed n log n bits copressed CSA = T BWT + SA saples: n log σ + o(n log σ) bits + n log n s SA bits

42 Introdction CST design CST in practice Exaple: Copressing CSA (practical approach) SA[13]=SA[9]+2 n l d l l n s SA = 3 access LF[i] in tie O(log σ) access CSA[i] in tie O(s SA log σ) l d SA ncopressed n log n bits copressed CSA = T BWT + SA saples: n log σ + o(n log σ) bits + n log n s SA bits

43 Introdction CST design CST in practice Exaple: Copressing CSA (practical approach) SA[13]=6+2=8 n l d l l n s SA = 3 access LF[i] in tie O(log σ) access CSA[i] in tie O(s SA log σ) l d SA ncopressed n log n bits copressed CSA = T BWT + SA saples: n log σ + o(n log σ) bits + n log n s SA bits

44 Introdction CST design CST in practice Overview of LCP data strctres data strctre ses access eory in bits lcp_ncopressed - O(1) n log n lcp_spport_sada CSA O(t SA ) 2n + o(n) lcp_krtz - O(log n) or O(1) 8n 1 + 2n 2 log n lcp_spport_tree NAV O(log σ ) H q 1 + q 2 log n lcp_spport_tree2 NAV & LF O(s LCP log σ ) H q 1 + (q 2 log n)/s LCP with n 1 + n 2 = n and q 1 + q 2 = q < n q nber of inner nodes of the ST

45 Introdction CST design CST in practice Rntie for rando access to the LCP array (1) dblp.xl.2mb proteins.2mb Tie in nanoseconds per operation lcp_ncopressed lcp_wt lcp_dac lcp_krtz lcp_spport_sada lcp_spport_tree lcp_spport_tree Meory in bits per character Meory in bits per character

46 Introdction CST design CST in practice Rntie for rando access to the LCP array (2) dna.2mb rand_k128.2mb Tie in nanoseconds per operation lcp_ncopressed lcp_wt lcp_dac lcp_krtz lcp_spport_sada lcp_spport_tree lcp_spport_tree Meory in bits per character Meory in bits per character

47 Introdction CST design CST in practice Rntie for rando access to the LCP array (3) english.2mb sorces.2mb Tie in nanoseconds per operation lcp_ncopressed lcp_wt lcp_dac lcp_krtz lcp_spport_sada lcp_spport_tree lcp_spport_tree Meory in bits per character Meory in bits per character

48 Introdction CST design CST in practice Otline 1 Introdction Basic data strctres The sffix tree 2 CST design NAV (tree topology and navigation) CSA (lexicographic inforation) LCP (longest coon prefixes) 3 CST in practice The sdsl library

49 Introdction CST design CST in practice The sccinct data strctre library sdsl Provides basic and advanced sccinct data strctres Easy to se (very siilar to C++ STL) Fast and space-efficient constrction of data strctres 64-bit ipleentation Well-optiized ipleentation (e.g. now sing hardware POPCOUNT operation,.. ) Easy configration of yriads of CSAs, CSTs with any tie-space trade-offs Fast prototyping of other coplex sccinct data strctres

50 Introdction CST design CST in practice cst_sada<csa_sada<>,lcp_.. cst_sct3<csa_wt<wt_hff<>.. vector has eber of class type vector has eber of class type int_vector int_vector enc_vector has teplate paraeter enc_vector has teplate paraeter rrr_vector of concept rrr_vector of concept rank_spport wavelet_tree select_spport rank_spport wavelet_tree select_spport rank_spport_v wt select_spport_cl rank_spport_v wt select_spport_cl rank_spport_v5 wt_int select_spport_bs rank_spport_v5 wt_int select_spport_bs rrr_rank_spport wt_hff rrr_select_spport rrr_rank_spport wt_hff rrr_select_spport wt_rln wt_rln wt_rlg wt_rlg csa csa csa_ncopressed csa_ncopressed csa_sada csa_sada csa_wt csa_wt lcp bp_spport lcp bp_spport lcp_ncopressed bp_spport_g lcp_ncopressed bp_spport_g lcp_dac bp_spport_gg lcp_dac bp_spport_gg lcp_wt bp_spport_sada lcp_wt bp_spport_sada lcp_krtz lcp_krtz lcp_spport_sada rq lcp_spport_sada rq lcp_spport_tree rq_spport_sparse_table lcp_spport_tree rq_spport_sparse_table lcp_spport_tree2 rq_sccinct_sct lcp_spport_tree2 rq_sccinct_sct rq_sccinct_sada rq_sccinct_sada cst cst cst_sada cst_sada cst_sct3 cst_sct3

51 Introdction CST design CST in practice CST space in practice (for english.2mb) MB 27% CSA: 18.8 MB wt: 148 MB data: 114 MB rank: 7.1 MB select 1: 14.4 MB select : 12.5 MB sa_saple: 21.9 MB isa_saple: 1.9 MB lcp: 215 MB lcp vales: 17.4 MB overflow ark: 41.9 MB rank: 2.6 MB nav: MB BPS dfs (bit_vector): 8.9 MB bp_spport: 37.1 MB sall block: 5.7 MB edi block: 1.6 MB bp rank: 2.2 MB bp select: 9.6 MB rank_spport1: 2.2 MB select_spport1: 6.1 MB cst_sada<csa_wt<wt_hff<> >,lcp_dac<> >

52 Introdction CST design CST in practice CST space in practice (for english.2mb) MB 23% 114. CSA: 18.8 MB wt: 148 MB data: 114 MB rank: 7.1 MB select 1: 14.4 MB select : 12.5 MB sa_saple: 21.9 MB isa_saple: 1.9 MB lcp: 96.1 MB sall lcp: 72.1 MB data: 67.9 MB rank: 4.2 MB big lcp: 24 MB nav: MB bp: 8.9 MB bp_spport: 21.9 MB sall block: 5.7 MB edi block: 1.6 MB bp rank: 5.1 MB bp select: 9.6 MB rank_spport1: 2.2 MB select_spport1: 6.1 MB cst_sada<csa_wt<wt_hff<> >,lcp_spport_tree2<> >

53 Introdction CST design CST in practice CST space in practice (for english.2mb) MB 167% CSA: 18.4 MB wt: 75.6 MB data: 75.6 MB bt: 3.4 MB btnr: 31.4 MB btnrp: 6.7 MB rank saples: 6.9 MB invert:.2 MB rank: MB select 1: MB select : MB sa_saple: 21.9 MB isa_saple: 1.9 MB lcp: 96.1 MB sall lcp: 72.1 MB data: 67.9 MB rank: 4.2 MB big lcp: 24 MB nav: MB BPS dfs (bit_vector): 8.9 MB bp_spport: 21.9 MB sall block: 5.7 MB edi block: 1.6 MB bp rank: 5.1 MB bp select: 9.6 MB rank_spport1: 2.2 MB select_spport1: 6.1 MB cst_sada<csa_wt<wt_hff<rrr_vector<> > >,lcp_spport_tree2<> >

54 Introdction CST design CST in practice CST space in practice (for english.2mb) MB 147% CSA: 18.4 MB wt: 75.6 MB data: 75.6 MB bt: 3.4 MB btnr: 31.4 MB btnrp: 6.7 MB rank saples: 6.9 MB invert:.2 MB rank: MB select 1: MB select : MB sa_saple: 21.9 MB isa_saple: 1.9 MB lcp: 96.1 MB sall lcp: 72.1 MB data: 67.9 MB rank: 4.2 MB big lcp: 24 MB nav: 88.4 MB bp: 5 MB bp_spport: 13.4 MB sall block: 3.5 MB edi block:.8 MB cst_sct3<csa_wt<wt_hff<rrr_vector<> > >,lcp_spport_tree2<> >

55 Introdction CST design CST in practice Experiental setp ˆ= cst_sada<csa_sada<>, lcp_dac<> > 1 ˆ= cst_sada<csa_sada<>, lcp_spport_tree2<> > 2 ˆ= cst_sada<csa_wt<>, lcp_dac<> > 3 ˆ= cst_sada<csa_wt<>, lcp_spport_tree2<> > 4 ˆ= cst_sct3<csa_sada<>, lcp_dac<> > 5 ˆ= cst_sct3<csa_sada<>, lcp_spport_tree2<> > 6 ˆ= cst_sct3<csa_wt<>, lcp_dac<> > 7 ˆ= cst_sct3<csa_wt<>, lcp_spport_tree2<> > The sae basic data strctres are sed, i.e. its a very fair coparison

56 Introdction CST design CST in practice Rntie of operations of cst_sada csa[i]: psi[i]: lcp[i]: id(v): depth(v): depth(v) : parent(v): sibling(v): sl(v): child(v, c): select child(v, 1): lca(v, w): dfs and id(v): dfs and depth(v): stats : µs 2µs 4µs 6µs µs 2µs 4µs 6µs

57 Introdction CST design CST in practice Rntie of operations of cst_sct3 csa[i]: psi[i]: lcp[i]: id(v): depth(v): depth(v) : parent(v): sibling(v): sl(v): child(v, c): select child(v, 1): lca(v, w): dfs and id(v): dfs and depth(v): stats : µs 2µs 4µs 6µs µs 2µs 4µs 6µs

58 Introdction CST design CST in practice Tie-space trade-off for select child(v, 1) english.2mb sorces.2mb Tie in nanoseconds per operation Meory in bits per character Meory in bits per character

59 Introdction CST design CST in practice Tie-space trade-off for child(v, c) english.2mb sorces.2mb Tie in nanoseconds per operation Meory in bits per character Meory in bits per character

60 Introdction CST design CST in practice Constrction of a CST #inclde < sdsl / s f f i x t r e e s. hpp> #inclde < sdsl / t i l. hpp> sing naespace sdsl ; typedef cst_sct3 <> tcst ; i n t ain ( i n t argc, char argv [ ] ) { tcst c s t ; c o n s t r c t _ c s t ( argv [ 1 ], c s t ) ; }

61 Introdction CST design CST in practice Rntie for constrction (prefixes of english text) 1 4 cstv ST cst_sada () cst_sct3 (4) Tie in seconds Text size in MB

62 Introdction CST design CST in practice Meory in bytes per character CST constrction resorces coparison text: english.2mb σ: 226 CST type: cst_sada LCP type: lcp_sada CSA type: csa_wt 5 first CST ipleentation (27) sdsl (21) NAV LCP 1 Tie in seconds CSA

63 Introdction CST design CST in practice Detailed resorces for the constrction of CSTs text : english.2mb CST type: cst_sada CSA type: csa_wt text : english.2mb CST type: cst_sada CSA type: csa_wt text : english.2mb CST type: cst_sct3 CSA type: csa_wt 1 Tie in seconds LCP: 67.2 CSA: 18.3 NAV: 9.9 wt: 9.9 SA: 57.4 bp_spport: 2.3 T BWT : 29.4 NAV: LCP: 65.9 CSA: 18.5 wt: 9.9 SA: 56.5 bp_spport: 23.5 T BWT : 29.6 NAV: LCP: 98.8 CSA: 16.8 wt: 9.9 SA: 56.7 bp_spport: 24.3 ISA: 33.9 T BWT : Meory in bytes per inpt character

64 Introdction CST design CST in practice Depth first search traversal in a CST teplate <class Cst > void t e s t _ c s t _ d f s _ i t e r a t o r _ a n d _ d e p t h ( Cst &c s t ) { typedef typenae Cst : : c o n s t _ i t e r a t o r i t e r a t o r ; long long cnt = ; for ( i t e r a t o r i t = c s t. begin ( ) ; i t!= c s t. end ( ) ; + + i t ) { i f (! c s t. i s _ l e a f ( i t ) ) cnt += c s t. depth ( i t ) ; } cot << cnt << endl ; }

65 Introdction CST design CST in practice Rntie of dfs on CST (prefixes of english text) 1 4 cstv ST cst_sada () cst_sct3 (4) Tie in nanoseconds Text size in MB

66 Introdction CST design CST in practice Conclsion CSTs can be bild fast and space-efficient... provide a rich set of fnctionality fast operations: basic navigation, access LF or Ψ slow operations: child(v, c), access to SA Yo can se the sdsl library to configre a CSTs which fits yor needs

67 Introdction CST design CST in practice Thank yo!

68 Introdction CST design CST in practice Rntie of int_vector access Tie in nanoseconds per operation rando access read rando access write seqential write int_vector<32> bit_vector int_vector<> v(..,..,32) int_vector<> v(..,..,27)

69 Introdction CST design CST in practice Operation rntie of basic data strctres Tie in nanoseconds per operation int_vector<32> rank_spport_v rank_spport_v5 select_spport_cl (rando access)

70 Introdction CST design CST in practice Rntie of CSA access (1) Tie in nanoseconds per operation dblp.xl.2mb proteins.2mb csa_wt<wt<>> csa_wt<wt_hff<>> csa_wt<wt_rln<>> csa_wt<wt_rlg<8>> csa_sada<δ> csa_sada<φ> Meory in bits per character Meory in bits per character

71 Introdction CST design CST in practice Rntie of CSA access (2) Tie in nanoseconds per operation dna.2mb rand_k128.2mb csa_wt<wt<>> csa_wt<wt_hff<>> csa_wt<wt_rln<>> csa_wt<wt_rlg<8>> csa_sada<δ> csa_sada<φ> Meory in bits per character Meory in bits per character

72 Introdction CST design CST in practice Rntie of CSA access (3) Tie in nanoseconds per operation english.2mb sorces.2mb csa_wt<wt<>> csa_wt<wt_hff<>> csa_wt<wt_rln<>> csa_wt<wt_rlg<8>> csa_sada<δ> csa_sada<φ> Meory in bits per character Meory in bits per character

73 Introdction CST design CST in practice Rntie of CSA operations (1) Tie in nanoseconds per operation dblp.xl.2mb proteins.2mb csa_wt<wt<>> csa_wt<wt_hff<>> csa_wt<wt_rln<>> csa_wt<wt_rlg<8>> csa_sada<δ> csa_sada<φ> Meory in bits per character Meory in bits per character =psi[i], =psi(i)=lf[i], + =bwt[i]

74 Introdction CST design CST in practice Rntie of CSA operations (2) Tie in nanoseconds per operation dna.2mb rand_k128.2mb csa_wt<wt<>> csa_wt<wt_hff<>> csa_wt<wt_rln<>> csa_wt<wt_rlg<8>> csa_sada<δ> csa_sada<φ> Meory in bits per character Meory in bits per character =psi[i], =psi(i)=lf[i], + =bwt[i]

75 Introdction CST design CST in practice Rntie of CSA operations (3) Tie in nanoseconds per operation english.2mb sorces.2mb csa_wt<wt<>> csa_wt<wt_hff<>> csa_wt<wt_rln<>> csa_wt<wt_rlg<8>> csa_sada<δ> csa_sada<φ> Meory in bits per character Meory in bits per character =psi[i], =psi(i)=lf[i], + =bwt[i]

MODELLING THE EFFECTS OF PEDESTRIANS ON INTERSECTION CAPACITY AND DELAY WITH ACTUATED SIGNAL CONTROL

MODELLING THE EFFECTS OF PEDESTRIANS ON INTERSECTION CAPACITY AND DELAY WITH ACTUATED SIGNAL CONTROL MODELLING THE EFFECTS OF PEDESTRIANS ON INTERSECTION CAPACITY AND DELAY WITH ACTUATED SIGNAL CONTROL ABSTRACT Zong Tian, Ph.D., P.E. Feng Xu, Graduate Assistant Departent of Civil and Environental Engineering

More information

Exercise 1 Network Planning Design. Teresa Grilo Dep. of Engineering and Management Instituto Superior Técnico Lisbon, Portugal

Exercise 1 Network Planning Design. Teresa Grilo Dep. of Engineering and Management Instituto Superior Técnico Lisbon, Portugal Exercise 1 Network Planning Design Teresa Grilo Dep. of Engineering and Manageent Instituto Superior Técnico Lisbon, Portugal SCM/IST 1 SCM/IST IST Lisbon Ana Ana Carvalho 2015 2015 Introduction Find the

More information

An improvement in calculation method for apparel assembly line balancing

An improvement in calculation method for apparel assembly line balancing Indian Journal of Fibre & Textile Research Vol.38, Septeber 2013, pp 259-264 An iproveent in calculation ethod for apparel assebly line balancing F Khosravi 1, a, A H Sadeghi 1 & F Jolai 2 1 Textile Departent,

More information

Guide to Target Archery Rounds, Scoring, Handicaps and Classifications

Guide to Target Archery Rounds, Scoring, Handicaps and Classifications Guide to Target Archery Rounds, Scoring, Handicaps and Classifications 2016 Contents 1Target Archery Rounds...3 1.1Iperial Outdoor Rounds...4 1.2Metric Outdoor Rounds...5 1.3Indoor Rounds...6 1.4Shooting

More information

Guide to Target Archery Rounds, Scoring, Handicaps and Classifications

Guide to Target Archery Rounds, Scoring, Handicaps and Classifications Guide to Target Archery Rounds, Scoring, Handicaps and Classifications Produced by Malcol North (Herts Group of Archery Coaches) Page 1 of 11 Contents 1Target Archery Rounds...3 1.1Iperial Outdoor Rounds...3

More information

Naïve Bayes Classifiers

Naïve Bayes Classifiers For a ore in-depth introdction to Naïe Bayes lassifiers and the theory srronding the, please see Andrew s lectre on robability for Data Miners. Naïe Bayes lassifiers Note to other teachers and sers of

More information

Sensor Technologie A TEMPOSONICS. R Series CANbasic C207. Operating Manual. Configuration and CAN-Bus Coupling. page PL.

Sensor Technologie A TEMPOSONICS. R Series CANbasic C207. Operating Manual. Configuration and CAN-Bus Coupling. page PL. Sensor Technologie TEMPOSONICS R 551040 A R Series 2004 CANbasic C207 Operating Manual Configuration and CAN-Bus Coupling page 1 Operating anual for CANbasic C207 Connecting diagra: RP-x-xxxxM-D6x-1-C207

More information

Natural gas consideration in a medium-term electricity market equilibrium model

Natural gas consideration in a medium-term electricity market equilibrium model ESCUELA TÉCNICA SUPERIOR DE INGENIERÍA ICAI Natral gas consideration in a medim-term electricity market eqilibrim model Pablo Deñas, Jlián Barqín, Javier Reneses Introdction Natral gas has been increasing

More information

Solving Problems by Searching chap3 1. Problem-Solving Agents

Solving Problems by Searching chap3 1. Problem-Solving Agents Chapter3 Solving Problems by Searching 20070315 chap3 1 Problem-Solving Agents Reflex agents cannot work well in those environments - state/action mapping too large - take too long to learn Problem-solving

More information

COMP219: Artificial Intelligence. Lecture 8: Combining Search Strategies and Speeding Up

COMP219: Artificial Intelligence. Lecture 8: Combining Search Strategies and Speeding Up COMP219: Artificial Intelligence Lecture 8: Combining Search Strategies and Speeding Up 1 Overview Last time Basic problem solving techniques: Breadth-first search complete but expensive Depth-first search

More information

Overview. Depth Limited Search. Depth Limited Search. COMP219: Artificial Intelligence. Lecture 8: Combining Search Strategies and Speeding Up

Overview. Depth Limited Search. Depth Limited Search. COMP219: Artificial Intelligence. Lecture 8: Combining Search Strategies and Speeding Up COMP219: Artificial Intelligence Lecture 8: Combining Search Strategies and Speeding Up Last time Basic problem solving techniques: Breadth-first search complete but expensive Depth-first search cheap

More information

Problem Solving Agents

Problem Solving Agents Problem Solving Agents A problem solving agent is one which decides what actions and states to consider in completing a goal Examples: Finding the shortest path from one city to another 8-puzzle Problem

More information

Flyweight Pattern. Flyweight: Intent. Use sharing to support large numbers of fine-grained objects efficiently. CSIE Department, NTUT Chien-Hung Liu

Flyweight Pattern. Flyweight: Intent. Use sharing to support large numbers of fine-grained objects efficiently. CSIE Department, NTUT Chien-Hung Liu Flyweight Pattern CSIE Department, NTUT Chien-Hung Liu Flyweight: Intent Use sharing to support large numbers of fine-grained objects efficiently 1 Flyweight: Motivation (1) Some applications could benefit

More information

Integration of Lean Approaches to Manage a Manual Assembly System

Integration of Lean Approaches to Manage a Manual Assembly System Open Journal of Social Sciences, 204, 2, 226-23 Published Online Septeber 204 in SciRes. http://www.scirp.org/journal/jss http://dx.doi.org/0.4236/jss.204.29038 Integration of Lean Approaches to Manage

More information

IMPROVED SPECTRAL WAVE MODELLING OF WHITE-CAPPING DISSIPATION IN SWELL SEA SYSTEMS

IMPROVED SPECTRAL WAVE MODELLING OF WHITE-CAPPING DISSIPATION IN SWELL SEA SYSTEMS Proceedings of OMAE 2004: 23rd International Conference on Offshore Mechanics and Arctic Engineering 20-25 June 2004, Vancouver, Canada OMAE2004-51562 IMPROVED SPECTRAL WAVE MODELLING OF WHITE-CAPPING

More information

The Handtmann Armaturenfabrik. Safety without compromise. Safety valves for liquids, gases, and steam

The Handtmann Armaturenfabrik. Safety without compromise. Safety valves for liquids, gases, and steam The Handtann Araturenfabrik Safety without coproise for liquids, gases, and stea f Safety Valves FoodSafe Tradition Meets Innovation fro Handtann are first choice for hygienically sensitive product areas

More information

Word-based Statistical Compressors as Natural Language Compression Boosters

Word-based Statistical Compressors as Natural Language Compression Boosters Word-based Statistical Compressors as Natural Language Compression Boosters Antonio Fariña 1, Gonzalo Navarro 2, and José R. Paramá 1 1 Database Lab, University of A Coruña, A Coruña, Spain. {fari,parama}@udc.es

More information

Introduction. AI and Searching. Simple Example. Simple Example. Now a Bit Harder. From Hammersmith to King s Cross

Introduction. AI and Searching. Simple Example. Simple Example. Now a Bit Harder. From Hammersmith to King s Cross Introduction AI and Searching We have seen how models of the environment allow an intelligent agent to dry run scenarios in its head without the need to act Logic allows premises to be tested Machine learning

More information

CSE 3401: Intro to AI & LP Uninformed Search II

CSE 3401: Intro to AI & LP Uninformed Search II CSE 3401: Intro to AI & LP Uninformed Search II Required Readings: R & N Chapter 3, Sec. 1-4. 1 {Arad}, {Zerind, Timisoara, Sibiu}, {Zerind, Timisoara, Arad, Oradea, Fagaras, RimnicuVilcea }, {Zerind,

More information

Satoshi Yoshida and Takuya Kida Graduate School of Information Science and Technology, Hokkaido University

Satoshi Yoshida and Takuya Kida Graduate School of Information Science and Technology, Hokkaido University Satoshi Yoshida and Takuya Kida Graduate School of Information Science and Technology, Hokkaido University ompressed Pattern Matching ompressed Data Search Directly 0000 000000 Program Searching on ompressed

More information

CENG 466 Artificial Intelligence. Lecture 4 Solving Problems by Searching (II)

CENG 466 Artificial Intelligence. Lecture 4 Solving Problems by Searching (II) CENG 466 Artificial Intelligence Lecture 4 Solving Problems by Searching (II) Topics Search Categories Breadth First Search Uniform Cost Search Depth First Search Depth Limited Search Iterative Deepening

More information

Preventive Strike vs. False Targets in Defense Strategy

Preventive Strike vs. False Targets in Defense Strategy International Journal of Perforability Engineering Vol. 8, No., July, pp. -. RAMS Consultants Printed in India Preventive Strike vs. False Targets in Defense Strategy GREGORY LEVITIN and KJELL HAUSKEN.

More information

Reducing Code Size with Run-time Decompression

Reducing Code Size with Run-time Decompression Reducing Code Size with Run-time Decompression Charles Lefurgy, Eva Piccininni, and Trevor Mudge Advanced Computer Architecture Laboratory Electrical Engineering and Computer Science Dept. The University

More information

Cardiac Science Trainer AED Manual

Cardiac Science Trainer AED Manual Cardiac Science Trainer AED Manal Get an original copy of the Cardiac Science Trainer AED Manal for manfactrer information abot service, available accessories and how to se and maintain yor device. AED

More information

Computing s Energy Problem:

Computing s Energy Problem: Computing s Energy Problem: (and what we can do about it) Mark Horowitz Stanford University horowitz@ee.stanford.edu 1 of 46 Everything Has A Computer Inside 2of 46 The Reason is Simple: Moore s Law Made

More information

Race car damping 2. Fig-1 quarter car model.

Race car damping 2. Fig-1 quarter car model. Race car daping A nuber of issues ago I wrote an article on exploring approaches to specify a race car daper. This article is the second in that series and we shall be following on fro where we left off.

More information

SUPPLEMENT MATERIALS

SUPPLEMENT MATERIALS SUPPLEMENT MATERIALS This document provides the implementation details of LW-FQZip 2 and the detailed experimental results of the comparison studies. 1. Implementation details of LW-FQZip 2 LW-FQZip 2

More information

Jurnal Teknologi A MODIFIED ARTIFICIAL POTENTIAL FIELD METHOD FOR RIVERINE OBSTACLES AVOIDANCE. Full Paper. Mei Jian Hong, M. R.

Jurnal Teknologi A MODIFIED ARTIFICIAL POTENTIAL FIELD METHOD FOR RIVERINE OBSTACLES AVOIDANCE. Full Paper. Mei Jian Hong, M. R. Jurnal Teknologi A MODIFIED ATIFICIAL POTENTIAL FIELD METHOD FO IVEINE OBSTACLES AVOIDANCE Mei Jian Hong, M.. Arshad* Underwater, Control and obotics esearch Group (UCG), School of Electrical and Electronic

More information

CT PET-2018 Part - B Phd COMPUTER APPLICATION Sample Question Paper

CT PET-2018 Part - B Phd COMPUTER APPLICATION Sample Question Paper CT PET-2018 Part - B Phd COMPUTER APPLICATION Sample Question Paper Note: All Questions are compulsory. Each question carry one mark. 1. Error detection at the data link layer is achieved by? [A] Bit stuffing

More information

The water outgassing rate of internal surfaces of vacuum systems

The water outgassing rate of internal surfaces of vacuum systems Journal of Physics: Conference Series PAPER OPEN ACCESS The water outgassing rate of internal surfaces of vacuu systes Recent citations - Soe probles of the vacuu technique developent L N Rozanov and A

More information

Better Search Improved Uninformed Search CIS 32

Better Search Improved Uninformed Search CIS 32 Better Search Improved Uninformed Search CIS 32 Functionally PROJECT 1: Lunar Lander Game - Demo + Concept - Open-Ended: No One Solution - Menu of Point Options - Get Started NOW!!! - Demo After Spring

More information

Lecture 7: Pipelined Implementation. James C. Hoe Department of ECE Carnegie Mellon University

Lecture 7: Pipelined Implementation. James C. Hoe Department of ECE Carnegie Mellon University 8 447 Lectre 7: Pipelined Implementation James C. Hoe Department of ECE Carnegie ellon University 8 447 S8 L7 S, James C. Hoe, CU/ECE/CALC, 28 Yor goal today 8 447 S8 L7 S2, James C. Hoe, CU/ECE/CALC,

More information

Problem Solving as Search - I

Problem Solving as Search - I Problem Solving as Search - I Shobhanjana Kalita Dept. of Computer Science & Engineering Tezpur University Slides prepared from Artificial Intelligence A Modern approach by Russell & Norvig Problem-Solving

More information

AN OPTIMIZATION MODEL AND ALGORITHM OF STUDENTS' PHYSICAL FITNESS TEST SEQUENCE

AN OPTIMIZATION MODEL AND ALGORITHM OF STUDENTS' PHYSICAL FITNESS TEST SEQUENCE 28 th February 203. Vol. 48 3 2005-203 JATIT & LLS. All rights reserved. ISSN: 992-8645 www.jatit.org E-ISSN: 87-395 AN OPTIMIZATION MODEL AND ALGORITHM OF STUDENTS' PHYSICAL FITNESS TEST SEQUENCE JINSONG

More information

Fast Software-managed Code Decompression

Fast Software-managed Code Decompression Fast Software-managed Code Decompression Charles Lefurgy and Trevor Mudge Advanced Computer Architecture Laboratory Electrical Engineering and Computer Science Dept. The University of Michigan, Ann Arbor

More information

GOLOMB Compression Technique For FPGA Configuration

GOLOMB Compression Technique For FPGA Configuration GOLOMB Compression Technique For FPGA Configuration P.Hema Assistant Professor,EEE Jay Shriram Group Of Institutions ABSTRACT Bit stream compression is important in reconfigurable system design since it

More information

Exploring fisheries dynamics according to different quota governance mechanisms: application to the bay of Biscay common sole fishery

Exploring fisheries dynamics according to different quota governance mechanisms: application to the bay of Biscay common sole fishery IIFET 2014 Conference, 7-11 July 2014, Brisbane, Australia Exploring fisheries dynaics according to different quota governance echaniss: application to the bay of Biscay coon sole fishery Manuel Bellanger,

More information

The system design must obey these constraints. The system is to have the minimum cost (capital plus operating) while meeting the constraints.

The system design must obey these constraints. The system is to have the minimum cost (capital plus operating) while meeting the constraints. Computer Algorithms in Systems Engineering Spring 2010 Problem Set 6: Building ventilation design (dynamic programming) Due: 12 noon, Wednesday, April 21, 2010 Problem statement Buildings require exhaust

More information

Overview: Underwater sensing

Overview: Underwater sensing Overview: Underwater sensing Vasilescu, I., Kotay, K., Rus, D., Dunbabin, M., and Corke, P. Data collection, storage, and retrieval with an underwater sensor network. In SenSys '05 Experimental results

More information

Uninformed search methods

Uninformed search methods Lecture 3 Uninformed search methods Milos Hauskrecht milos@cs.pitt.edu 5329 Sennott Square Announcements Homework 1 Access through the course web page http://www.cs.pitt.edu/~milos/courses/cs2710/ Two

More information

Uninformed search methods

Uninformed search methods Lecture 3 Uninformed search methods Milos Hauskrecht milos@cs.pitt.edu 5329 Sennott Square Announcements Homework assignment 1 is out Due on Tuesday, September 12, 2017 before the lecture Report and programming

More information

Decompression Method For Massive Compressed Files In Mobile Rich Media Applications

Decompression Method For Massive Compressed Files In Mobile Rich Media Applications 2010 10th IEEE International Conference on Computer and Information Technology (CIT 2010) Decompression Method For Massive Compressed Files In Mobile Rich Media Applications Houchen Li, Zhijie Qiu, Lei

More information

Design Strategies for ARX with Provable Bounds: SPARX and LAX

Design Strategies for ARX with Provable Bounds: SPARX and LAX Design Strategies for ARX with Provable Bounds: SPARX and LAX Daniel Dinu 1, Léo Perrin 1, Aleksei Udovenko 1, Vesselin Velichkov 1, Johann Großschädl 1, Alex Biryukov 1 1 SnT, University of Luxembourg

More information

A Numerical Prediction of Wash Wave and Wave Resistance of High Speed Displacement Ships in Deep and Shallow Water

A Numerical Prediction of Wash Wave and Wave Resistance of High Speed Displacement Ships in Deep and Shallow Water การประช มว ชาการเคร อข ายว ศวกรรมเคร องกลแห งประเทศไทยคร งท 18 18- ต ลาคม 547 จ งหว ดขอนแก น A Nuerical Prediction of Wash Wave and Wave Resistance of High Speed Displaceent Ships in Deep and Shallow Water

More information

PC-based systems ELOP II-NT

PC-based systems ELOP II-NT ELOP II-NT... the safe decision. 1 (mm 9910) Continuity in the work process... Double work processes resulting from a lack of interfaces between project stages is an unnecessary waste of time and money

More information

CSE 3402: Intro to Artificial Intelligence Uninformed Search II

CSE 3402: Intro to Artificial Intelligence Uninformed Search II CSE 3402: Intro to Artificial Intelligence Uninformed Search II Required Readings: Chapter 3, Sec. 1-4. 1 {Arad}, {Zerind, Timisoara, Sibiu}, {Zerind, Timisoara, Arad, Oradea, Fagaras, RimnicuVilcea },

More information

Weight Optimization and Finite Element Analysis of Pressure Vessel Due to Thickness

Weight Optimization and Finite Element Analysis of Pressure Vessel Due to Thickness 2017 IJSRSET Volue 3 Issue 2 Print ISSN: 2395-1990 Online ISSN : 2394-4099 Theed Section: Engineering and Technology Weight Optiization and Finite Eleent Analysis of Pressure Vessel Due to Thickness Rashi

More information

Tabulation of Bids. Counties: Project(s):

Tabulation of Bids. Counties: Project(s): Page: 1 14 ROTE 94, ROTE 46 TO OLD COACH RD Quantity and nits nit Price nit Price nit Price 0001 151006M PERFORMANCE BOND AND PAYMENT BOND 0004 154003P MOBILIZATION 0007 157004M CONSTRCTION LAYOT 0012

More information

Name: Answer Key Date: Regents Physics. Waves

Name: Answer Key Date: Regents Physics. Waves Nae: Answer Key Date: Regents Physics Test # 13 Review Waves 1. Use GUESS ethod and indicate all vector directions. 2. Ters to know: echanical wave, transverse wave, longitudinal wave, surace wave, electroagnetic

More information

Fast Floating Point Compression on the Cell BE Processor

Fast Floating Point Compression on the Cell BE Processor Fast Floating Point Compression on the Cell BE Processor Ajith Padyana, T.V. Siva Kumar, P.K.Baruah Sri Satya Sai University Prasanthi Nilayam - 515134 Andhra Pradhesh, India ajith.padyana@gmail.com, tvsivakumar@gmail.com,

More information

The Economic Contributions of Winter Sports in a Changing Climate Marca Hagenstad. January 2018

The Economic Contributions of Winter Sports in a Changing Climate Marca Hagenstad. January 2018 The Economic Contribtions of Winter Sports in a Changing Climate Marca Hagenstad Janary 2018 Contents Stdy Backgrond Participation in Winter Sports Economics Expenditres Economic Modeling Contribtions

More information

COMPETITION RULES FOR DOUBLE AXE THROWING Germany / Switzerland

COMPETITION RULES FOR DOUBLE AXE THROWING Germany / Switzerland COMPETITION RULES FOR DOUBLE AXE THROWING Germany / Switzerland Version 2016 Double Axe (6.10 meter or 20 feet) Two-Handed thrown at targets with 5cm/ or 1,96 inch rings After 10 years of experience and

More information

Kerk Lead-screw Assemblies: Lead-Screws Ø 1/8 to 15/16-in (3.2 to 23 mm)

Kerk Lead-screw Assemblies: Lead-Screws Ø 1/8 to 15/16-in (3.2 to 23 mm) Kerk -screw Assebies: -Screws Ø 1/8 to 15/16-in (3.2 to 23 ) Kerk -screws are anufactured fro 303 stainess stee and are produced with Kerk s excusive precision roing process. Kerk ead-screws are avaiabe

More information

Simulating the effects of fatigue and pathologies in gait

Simulating the effects of fatigue and pathologies in gait Siulating the effects of fatigue and pathologies in gait T Koura 1, A Nagano, H Leung 3 and Y Shinagawa 4 1,3 Departent of Coputer Engineering and Inforation Technology, City University of Hong Kong, 83

More information

Estimation of the Effect of Cadence on Gait Stability in Young and Elderly People using Approximate Entropy Technique

Estimation of the Effect of Cadence on Gait Stability in Young and Elderly People using Approximate Entropy Technique MEASUREMENT SCIENCE REVIEW, Volue 4, Section 2, 4 Estiation of the Effect of Cadence on Gait Stability in Young and Elderly People using Approxiate Entropy Technique M. Arif *, Y. Ohtaki **, R. Nagatoi

More information

Ratio Deco. Overview. Global Perspective

Ratio Deco. Overview. Global Perspective Ratio Deco Overview Ratio deco is simply an on the fly system of applying a set of rules to develop a decompression strategy that will work for you in your range of diving. These rules are derived from

More information

Algorithms and Data Structures

Algorithms and Data Structures Algorithms and Data Structures CMPSC 465 LECTURES 22-23 Binary Search Trees Adam Smith S. Raskhodnikova and A. Smith. Based on slides by C. Leiserson and E. Demaine. 1 Heaps: Review Heap-leap-jeep-creep(A):

More information

Hierarchical ORAM Revisited, and Applications to Asymptotically Efficient ORAM and OPRAM. Hubert Chan, Yue Guo, Wei-Kai Lin, Elaine Shi 2017/12/5

Hierarchical ORAM Revisited, and Applications to Asymptotically Efficient ORAM and OPRAM. Hubert Chan, Yue Guo, Wei-Kai Lin, Elaine Shi 2017/12/5 Hierarchical ORAM Revisited, and Applications to Asymptotically Efficient ORAM and OPRAM Hubert Chan, Yue Guo, Wei-Kai Lin, Elaine Shi 2017/12/5 Random Access Machine, RAM Maybe the standard model of algorithms

More information

AccuRAID iscsi Auto-Tiering Best Practice

AccuRAID iscsi Auto-Tiering Best Practice AccuRAID iscsi Auto-Tiering Best Practice Overview From the perspective of storage features, the performance of SSDs are high, but the cost is also high per GB. Relatively speaking, the cost of a traditional

More information

2013 CIVL PLENARY ANNEX 24B SOFTWARE PROPOSALS

2013 CIVL PLENARY ANNEX 24B SOFTWARE PROPOSALS 2013 CIVL PLENARY ANNEX 24B SOFTWARE PROPOSALS In conjunction with the Hang Gliding, Paragliding and Aerobatic Subcommittees 1. Speed rank used to calculate time validity At the 2008 CIVL plenary, the

More information

Princess Nora University Faculty of Computer & Information Systems ARTIFICIAL INTELLIGENCE (CS 370D) Computer Science Department

Princess Nora University Faculty of Computer & Information Systems ARTIFICIAL INTELLIGENCE (CS 370D) Computer Science Department Princess Nora University Faculty of Computer & Information Systems 1 ARTIFICIAL INTELLIGENCE (CS 370D) Computer Science Department (CHAPTER-3-PART2) PROBLEM SOLVING AND SEARCH (Course coordinator) Searching

More information

Basic Instrument Scan. T6BDriver.com Created: 4 Feb 2016 Updated: 28 Aug 2016

Basic Instrument Scan. T6BDriver.com Created: 4 Feb 2016 Updated: 28 Aug 2016 Basic Instrument Scan T6BDriver.com Created: 4 Feb 2016 Updated: 28 Aug 2016 Information Sources Attitude Instrument Flying Control-Performance Method Scanning (Cross-check) Scanning Errors Scan Pattern

More information

Basics of Flow measurement using Hot-film anemometer

Basics of Flow measurement using Hot-film anemometer Basics of Flow easureent using Hot-fil aneoeter Version 1.0 1/9 Inhaltsverzeichnis: 1. Definitions 1.1. Air Velocity 1.2. Aount (of Gas) 1.. Flow 1..1. Mass flow rate 1..2. Voluetric flow rate 1... Standard

More information

CS Lecture 5. Vidroha debroy. Material adapted courtesy of Prof. Xiangnan Kong and Prof. Carolina Ruiz at Worcester Polytechnic Institute

CS Lecture 5. Vidroha debroy. Material adapted courtesy of Prof. Xiangnan Kong and Prof. Carolina Ruiz at Worcester Polytechnic Institute CS 3353 Lecture 5 Vidroha debroy Material adapted courtesy of Prof. Xiangnan Kong and Prof. Carolina Ruiz at Worcester Polytechnic Institute Searching for a number Find a specific number 91-3 26 54 73

More information

Queue analysis for the toll station of the Öresund fixed link. Pontus Matstoms *

Queue analysis for the toll station of the Öresund fixed link. Pontus Matstoms * Queue analysis for the toll station of the Öresund fixed link Pontus Matstoms * Abstract A new simulation model for queue and capacity analysis of a toll station is presented. The model and its software

More information

Heap Sort. Lecture 35. Robb T. Koether. Hampden-Sydney College. Mon, Apr 25, 2016

Heap Sort. Lecture 35. Robb T. Koether. Hampden-Sydney College. Mon, Apr 25, 2016 Heap Sort Lecture 35 Robb T. Koether Hampden-Sydney College Mon, Apr 25, 2016 Robb T. Koether (Hampden-Sydney College) Heap Sort Mon, Apr 25, 2016 1 / 14 1 Sorting 2 The Heap Sort Robb T. Koether (Hampden-Sydney

More information

Persistent Memory Performance Benchmarking & Comparison. Eden Kim, Calypso Systems, Inc. John Kim, Mellanox Technologies, Inc.

Persistent Memory Performance Benchmarking & Comparison. Eden Kim, Calypso Systems, Inc. John Kim, Mellanox Technologies, Inc. Persistent Memory Performance Benchmarking & Comparison Eden Kim, Calypso Systems, Inc. John Kim, Mellanox Technologies, Inc. PM Benchmarking & Comparison Part 1: Test Plan & Workloads Eden Kim of Calypso

More information

Tresca s or Mises Yield Condition in Pressure Vessel Design

Tresca s or Mises Yield Condition in Pressure Vessel Design Tresca s or Mises Yield Condition in Pressure Vessel Design Franz Rauscher Institute for Pressure Vessels and Plant Technology Vienna University of Technology Austria Why using Tresca s yield condition

More information

Season Ticket Basketball 2003

Season Ticket Basketball 2003 Season Ticket Basketball 2003 Season Ticket Basketball 2003 is a text-based basketball simulation modeled after the world of professional basketball. It allows you to take control of a basketball team

More information

A history-based estimation for LHCb job requirements

A history-based estimation for LHCb job requirements A history-based estimation for LHCb job requirements Nathalie Rauschmayr on behalf of LHCb Computing 13th April 2015 Introduction N. Rauschmayr 2 How long will a job run and how much memory might it need?

More information

COMP 406 Lecture 05. Artificial. Fiona Yan Liu Department of Computing The Hong Kong Polytechnic University

COMP 406 Lecture 05. Artificial. Fiona Yan Liu Department of Computing The Hong Kong Polytechnic University COMP 406 Lecture 05 Artificial Intelligence Fiona Yan Liu Department of Computing The Hong Kong Polytechnic University Learning Outcomes of Intelligent Agents Search agents Uninformed search Informed search

More information

Representation. Representation. Representation. Representation. 8 puzzle.

Representation. Representation. Representation. Representation. 8 puzzle. You have a 3-liter jug and a 4-liter jug, and need to measure out exactly two liters of water. the jugs do not have any measurement lines you don't have any additional containers 8 puzzle. Measuring water.

More information

A Fair Target Score Calculation Method for Reduced-Over One day and T20 International Cricket Matches

A Fair Target Score Calculation Method for Reduced-Over One day and T20 International Cricket Matches A Fair Target Score Calculation Method for Reduced-Over One day and T20 International Cricket Matches Rohan de Silva, PhD. Abstract In one day internationals and T20 cricket games, the par score is defined

More information

Real World Search Problems. CS 331: Artificial Intelligence Uninformed Search. Simpler Search Problems. Example: Oregon. Search Problem Formulation

Real World Search Problems. CS 331: Artificial Intelligence Uninformed Search. Simpler Search Problems. Example: Oregon. Search Problem Formulation Real World Search Problems S 331: rtificial Intelligence Uninformed Search 1 2 Simpler Search Problems ssumptions bout Our Environment Fully Observable Deterministic Sequential Static Discrete Single-agent

More information

An Efficient Code Compression Technique using Application-Aware Bitmask and Dictionary Selection Methods

An Efficient Code Compression Technique using Application-Aware Bitmask and Dictionary Selection Methods An Efficient Code Compression Technique using Application-Aware Bitmask and Selection Methods Seok-Won Seong sseong@cise.ufl.edu Prabhat Mishra prabhat@cise.ufl.edu Department of Computer and Information

More information

PROJECT and MASTER THESES 2016/2017

PROJECT and MASTER THESES 2016/2017 PROJECT and MASTER THESES 2016/2017 Below you ll find proposed topics for project and master theses. Most of the proposed topics are just sketches. The detailed topics will be made in discussion between

More information

Grade: 8. Author(s): Hope Phillips

Grade: 8. Author(s): Hope Phillips Title: Tying Knots: An Introductory Activity for Writing Equations in Slope-Intercept Form Prior Knowledge Needed: Grade: 8 Author(s): Hope Phillips BIG Idea: Linear Equations how to analyze data from

More information

NUMERICAL STUDY OF WAVE-CURRENT INTERACTION USING HIGH RESOLUTION COUPLED MODEL IN THE KUROSHIO REGION

NUMERICAL STUDY OF WAVE-CURRENT INTERACTION USING HIGH RESOLUTION COUPLED MODEL IN THE KUROSHIO REGION NUMERICAL STUDY OF WAVE-CURRENT INTERACTION USING HIGH RESOLUTION COUPLED MODEL IN THE KUROSHIO REGION Hitoshi TAMURA Frontier Research Center for Global Change/JAMSTEC 373- Showaachi, Kanazawa-ku, Yokohaa,

More information

1. Functional description. Application program usage. 1.1 General. 1.2 Behavior on bus voltage loss and bus voltage. 1.

1. Functional description. Application program usage. 1.1 General. 1.2 Behavior on bus voltage loss and bus voltage. 1. Application program usage product family: Product type: Manufacturer: Name: Order-No.: Valve actuators Constant valve actuator Siemens Valve actuator AP 562/02 5WG1 562-7AB02 Commissioning For commissioning

More information

Recommendations on Two Acceleration Measurements with Low Strain Integrity Test

Recommendations on Two Acceleration Measurements with Low Strain Integrity Test Recoendations on Two Acceleration Measureents with Low Strain Integrity Test Liqun Liang, PhD., P.E., 1 Scott Webster, P.E., 2 and Marty Bixler, P.E. 3 1 Pile Dynaics Inc., 3725 Aurora Rd, Cleveland, Ohio

More information

SEISMIC SURVEY GREENLAND 2014 Underwater sound propagation for North East Greenland offshore seismic survey

SEISMIC SURVEY GREENLAND 2014 Underwater sound propagation for North East Greenland offshore seismic survey TGS February 2014 SEISMIC SURVEY GREENLAND 2014 Underwater sound propagation for North East Greenland offshore seismic survey Appendix: NEG14 modelling results Mark Mikaelsen PROJECT Seismic Survey Greenland

More information

Mechanical Waves and Sound

Mechanical Waves and Sound Mechanical Waves and Sound Mechanical Wave Medium Crest Trough Transverse wave Compression Rarefaction Longitudinal wave Surface wave Some Vocab to Know What are Mechanical Waves? Mechanical wave: disturbance

More information

Tabulation of Bids. Counties: ROUTE 202,ALEXAUKEN CREEK ROAD TO RT 179/CR514

Tabulation of Bids. Counties: ROUTE 202,ALEXAUKEN CREEK ROAD TO RT 179/CR514 Page 1 of 17 18 HNTERDON ROTE 202,ALEXAKEN CREEK ROAD TO RT 179/CR514 Quantity and nits nit Price nit Price nit Price SECTION: 0001 ROADWAY 0001 151006M PERFORMANCE BOND AND PAYMENT BOND 0004 153003P PROGRESS

More information

Extensive Games with Perfect Information

Extensive Games with Perfect Information Extensive Games with Perfect Information A strategic situation in which players play a game over time. At the point of making the decision, players are informed about the action chosen before by the others.

More information

Introduction to Parallelism in CASTEP

Introduction to Parallelism in CASTEP to ism in CASTEP Phil Hasnip August 2016 Recap: what does CASTEP do? CASTEP solves the Kohn-Sham equations for electrons in a periodic array of nuclei: Ĥ k [ρ]ψ bk = E bk ψ bk where particle b has the

More information

CS472 Foundations of Artificial Intelligence. Final Exam December 19, :30pm

CS472 Foundations of Artificial Intelligence. Final Exam December 19, :30pm CS472 Foundations of Artificial Intelligence Final Exam December 19, 2003 12-2:30pm Name: (Q exam takers should write their Number instead!!!) Instructions: You have 2.5 hours to complete this exam. The

More information

David C. Yen Miami University

David C. Yen Miami University The Information System Consultant's Handbook Systems Analysis and Design Miami UniversityVy and David C. Yen Miami University CRC Press Boca Raton London New York Washington, D.C. Contents Part I: Principles

More information

Designing Mechanisms for Reliable Internet-based Computing

Designing Mechanisms for Reliable Internet-based Computing Designing Mechanisms for Reliable Internet-based Computing Chryssis Georgiou University of Cyprus, Cyprus Antonio Fernandez Miguel Mosteiro Universidad Rey Juan Carlos, Spain Trustworthy Network Computing

More information

Transposition Table, History Heuristic, and other Search Enhancements

Transposition Table, History Heuristic, and other Search Enhancements Transposition Table, History Heuristic, and other Search Enhancements Tsan-sheng Hsu tshsu@iis.sinica.edu.tw http://www.iis.sinica.edu.tw/~tshsu 1 Abstract Introduce heuristics for improving the efficiency

More information

SIDDHARTH INSTITUTE OF ENGINEERING & TECHNOLOGY :: PUTTUR (AUTONOMOUS) Siddharth Nagar, Narayanavanam Road QUESTION BANK (DESCRIPTIVE)

SIDDHARTH INSTITUTE OF ENGINEERING & TECHNOLOGY :: PUTTUR (AUTONOMOUS) Siddharth Nagar, Narayanavanam Road QUESTION BANK (DESCRIPTIVE) Subject with Code : Data Structures(16MC806) Course & Specialization: MCA UNIT I Sorting, Searching and Directories 1. Explain how to sort the elements by using insertion sort and derive time complexity

More information

Track B: Particle Methods Part 2 PRACE Spring School 2012

Track B: Particle Methods Part 2 PRACE Spring School 2012 Track B: Particle Methods Part 2 PRACE Spring School 2012 Maciej Cytowski (ICM UW) PART2: Load-balancing and migration Zoltan set-up ~ 15 min. Load-balancing in Zoltan ~ 15 min. Hands-on exercises ~ 20

More information

2018 COMPETITION PACK

2018 COMPETITION PACK 2018 COMPETITION PACK So mch more than a competition... it's an experience! www.cheerbrandz.com WHY CHEERBRANDZ? Astralasia's Leading Cheer & Dance Event Prodcer Delivering the most dynamic, entertaining

More information

Operating Instructions

Operating Instructions Table of Contents Table of Contents... 2 Conventions Used in this Manual... 4 1. Introduction...5 1.1. Notes on this anual... 6 1.2. Modes Covered by this Manual... 6 2. Basic Operation...7 2.1. Turning

More information

CRICKET ONTOLOGY. Project Description :- Instructor :- Prof: Navjyothi Singh

CRICKET ONTOLOGY. Project Description :- Instructor :- Prof: Navjyothi Singh Instructor :- Prof: Navjyothi Singh CRICKET ONTOLOGY Project Team :- PV Sai Krishna (200402036) Abhilash I (200501004) Phani Chaitanya (200501076) Kranthi Reddy (200502008) Vidyadhar Rao (200601100) Project

More information

A New Searchable Variable-to-Variable Compressor

A New Searchable Variable-to-Variable Compressor A New Searchable Variable-to-Variable Compressor Nieves R. Brisaboa 1,AntonioFariña 1,JuanR.López 1 Gonzalo Navarro 2, and Eduardo R. López 1 1 Database Lab, University of A Coruña, A Coruña, Spain. {brisaboa,fari,mon,erodriguezl}@udc.es

More information

Design of Saddle Support for Horizontal Pressure Vessel Vinod Kumar, Navin Kumar, Surjit Angra, Prince Sharma

Design of Saddle Support for Horizontal Pressure Vessel Vinod Kumar, Navin Kumar, Surjit Angra, Prince Sharma World Acadey of Science, Engineering and Technology Vol:8, No:, 0 Design of Saddle Support for Horizontal Pressure Vessel Vinod Kuar, Navin Kuar, Surjit Angra, Prince Shara International Science Index,

More information

Fish Habitat Design, Operation and Reclamation Worksheets for

Fish Habitat Design, Operation and Reclamation Worksheets for Fish Habitat Design, Operation and Reclaation Worksheets for Appendix A Project Location Worksheet Applicant Nae: Strea Nae: Watershed Nae: Fish Habitat Suitability Classification Water Quality Zone Operation

More information

Aryeh Rappaport Avinoam Meir. Schedule automation

Aryeh Rappaport Avinoam Meir. Schedule automation Aryeh Rappaport Avinoam Meir Schedule automation Introduction In this project we tried to automate scheduling. Which is to help a student to pick the time he takes the courses so they want collide with

More information

Projectile Motion Lab (2019)

Projectile Motion Lab (2019) Nae: Date: Partner(s): Period: Projectile Motion Lab (2019) Object: Measure the velocity of a ball using two Photogates and coputer software for tiing. Apply concepts fro two-diensional kineatics to predict

More information

MEMORY is one of the key driving factors in embeddedsystem

MEMORY is one of the key driving factors in embeddedsystem IEEE TRANSACTIONS ON COMPUTER-AIDED DESIGN OF INTEGRATED CIRCUITS AND SYSTEMS, VOL. 27, NO. 4, APRIL 2008 673 Bitmask-Based Code Compression for Embedded Systems Seok-Won Seong and Prabhat Mishra, Member,

More information