Mercurial > hg > audiodb
comparison docs/spec/spec.tex @ 562:dfeb5ef768da
Usually a commitment phobe but everybody has they're first time. Include mant structural changes and additional definition in line with what Christophe has asked for. Some of the schemas are more complicated than Id like - I will try to think of more elegant ways of specification.
author | mas02md |
---|---|
date | Fri, 17 Apr 2009 16:52:15 +0000 |
parents | 5a248cedd3e9 |
children | 365d3c500ef3 |
comparison
equal
deleted
inserted
replaced
557:d3c96cbb91e3 | 562:dfeb5ef768da |
---|---|
9 \author{Mark d'Inverno and Christophe Rhodes} | 9 \author{Mark d'Inverno and Christophe Rhodes} |
10 | 10 |
11 | 11 |
12 \begin{document} | 12 \begin{document} |
13 \maketitle | 13 \maketitle |
14 | |
15 | 14 |
16 %%\begin{gendef}[X] | 15 %%\begin{gendef}[X] |
17 %% concat : (\seq (\seq X)) \fun (\seq X) | 16 %% concat : (\seq (\seq X)) \fun (\seq X) |
18 %%\where | 17 %%\where |
19 %% \forall xs : \seq X; xss : \seq (\seq X) @ \\ | 18 %% \forall xs : \seq X; xss : \seq (\seq X) @ \\ |
30 %% \t2 (\langle (0 \upto n-1) \dres xs \rangle) \cat (nfront~n~(tail~xs)) | 29 %% \t2 (\langle (0 \upto n-1) \dres xs \rangle) \cat (nfront~n~(tail~xs)) |
31 %% \end{gendef} | 30 %% \end{gendef} |
32 | 31 |
33 \section{The System Instance} | 32 \section{The System Instance} |
34 | 33 |
34 \newcommand{\LET}{\mathrel{\sf Let}} | |
35 | |
35 \newcommand{\mylet}{\methrel{\sf Let}} | 36 \newcommand{\mylet}{\methrel{\sf Let}} |
36 \newcommand{\FV}{\mathrel{~FV}} | 37 \newcommand{\FV}{\mathrel{~FV}} |
37 \newcommand{\V}{\mathrel{~FV^{d}}} | 38 \newcommand{\V}{\mathrel{~FV^{d}}} |
38 \newcommand{\U}{\mathrel{~FV^{1}}} | 39 \newcommand{\U}{\mathrel{~FV^{1}}} |
39 \newcommand{\R}{\mathrel{~R}} | 40 \newcommand{\R}{\mathrel{~R}} |
61 %% \end{zed} | 62 %% \end{zed} |
62 | 63 |
63 %% \begin{axdef} | 64 %% \begin{axdef} |
64 %% \mylog : \R \fun \R | 65 %% \mylog : \R \fun \R |
65 %% \end{axdef} | 66 %% \end{axdef} |
66 | |
67 | |
68 | |
69 | 67 |
70 % For every track it is possible to determine the lenth. | 68 % For every track it is possible to determine the lenth. |
71 | 69 |
72 % \begin{axdef} | 70 % \begin{axdef} |
73 % lenth : Track \fun Time | 71 % lenth : Track \fun Time |
116 \where | 114 \where |
117 tracks = \{ cat : collection; t : Track | t \in (\ran cat) @ t \} \\ | 115 tracks = \{ cat : collection; t : Track | t \in (\ran cat) @ t \} \\ |
118 \dom name = tracks | 116 \dom name = tracks |
119 \end{schema} | 117 \end{schema} |
120 | 118 |
121 | |
122 | |
123 | |
124 \item \textsf{Time} - the set of non-negative reals. | 119 \item \textsf{Time} - the set of non-negative reals. |
125 | 120 |
126 | 121 |
127 \begin{zed} | 122 \begin{zed} |
128 Time == \R | 123 Time == \R |
129 \end{zed} | 124 \end{zed} |
130 | |
131 | |
132 | |
133 | 125 |
134 | 126 |
135 \item \textsf{Interval} - defined as a continuous set of real numbers, represented as an ordered pair of reals with the second of the pair strictly greater. | 127 \item \textsf{Interval} - defined as a continuous set of real numbers, represented as an ordered pair of reals with the second of the pair strictly greater. |
136 | 128 |
137 \begin{flushright} | 129 \begin{flushright} |
205 \draw[->] (1.4,0.9) -- (1.5,0.2); | 197 \draw[->] (1.4,0.9) -- (1.5,0.2); |
206 \draw[o->] (-0.5,0.2) -- (2,0.2) node[anchor=north west] {\small $\mathbb{R}$}; | 198 \draw[o->] (-0.5,0.2) -- (2,0.2) node[anchor=north west] {\small $\mathbb{R}$}; |
207 \end{tikzpicture} | 199 \end{tikzpicture} |
208 \end{flushright} | 200 \end{flushright} |
209 | 201 |
210 %% \begin{zed} | 202 %%unchecked |
211 %% ContinuousIntervalIndex == IntervalIndex | 203 \begin{zed} |
204 ContIntIndex == \\ | |
205 \t1 \{ ii : IntervalIndex; i_j, i_{j+1}: Interval | \\ | |
206 \t3 \langle i_j, i_{j+1} \rangle \inseq ii \implies second ~ i_j \geq first ~ i_{j+1}@ ii \} | |
207 \end{zed} | |
208 | |
209 % In fact this enables us to specify such a data structure simply as a sequence of increasing times as we can assume the track starts at 0. | |
210 | |
211 % \begin{zed} | |
212 % ContIntIndex == \iseq Time \\ \\ \\ | |
213 % \end{zed} | |
214 | |
215 %% \begin{zed} | |
216 %% ContIntIndex == IntervalIndex | |
212 %% \end{zed} | 217 %% \end{zed} |
213 | |
214 %%unchecked | |
215 \begin{zed} | |
216 ContinuousIntervalIndex == \\ | |
217 \t1 \{ ii : IntervalIndex; i_j, | |
218 i_{j+1}: Interval | \\ | |
219 \t3 \langle i_j, i_{j+1} \rangle \inseq ii \implies second ~ i_j \geq first ~ i_{j+1}@ ii \} | |
220 \end{zed} | |
221 | |
222 | 218 |
223 \item A \textsf{Duration} is an amount of time. | 219 \item A \textsf{Duration} is an amount of time. |
224 | 220 |
225 \begin{zed} | 221 \begin{zed} |
226 Duration == Time | 222 Duration == Time |
239 \where | 235 \where |
240 \forall i : Interval @ intervalduration ~ i = second ~ i - first ~ i | 236 \forall i : Interval @ intervalduration ~ i = second ~ i - first ~ i |
241 \end{axdef} | 237 \end{axdef} |
242 | 238 |
243 | 239 |
244 \item \textsf{Index Duration} is the duration of a Continuous Interval Index. It is calculated by subtracting the first element of the first interval from the second element of the last interval. | 240 \item \textsf{Index Duration} is the duration of a Continuous Interval Index which is simply the second element of the last interval pair. |
245 | |
246 | 241 |
247 \begin{axdef} | 242 \begin{axdef} |
248 indexduration : ContinuousIntervalIndex \fun Time | 243 indexduration : ContIntIndex \fun Time |
249 \where | 244 \where |
250 \forall cii : ContinuousIntervalIndex @ \\ | 245 \forall cii : ContIntIndex @ indexduration~cii = first (last~cii) |
251 \t1 indexduration~cii = second (last~cii) - first (head~cii) | |
252 \end{axdef} | 246 \end{axdef} |
253 | 247 |
254 \item \textsf{Segmenter} - a process which computes an interval index for any track, represented as a function which maps any track to an interval index. | 248 \item \textsf{Segmenter} - a process which computes an interval index for any track, represented as a function which maps any track to an interval index. |
255 | 249 |
256 \begin{flushright} | 250 \begin{flushright} |
258 \input interval-index.tikz | 252 \input interval-index.tikz |
259 \input segmented-track.tikz | 253 \input segmented-track.tikz |
260 \end{tikzpicture} | 254 \end{tikzpicture} |
261 \end{flushright} | 255 \end{flushright} |
262 | 256 |
263 \begin{zed} | 257 |
264 Segmenter == Track \fun IntervalIndex \\ | 258 In this specification we will consider segmenters that produce a continuous interval index. |
259 | |
260 \begin{zed} | |
261 Segmenter == Track \fun ContIntIndex \\ | |
265 \end{zed} | 262 \end{zed} |
266 | 263 |
267 \item The \textsf{Length} of an interval index is the number of intervals contained within it. | 264 \item The \textsf{Length} of an interval index is the number of intervals contained within it. |
268 | 265 |
269 \begin{axdef} | 266 \begin{axdef} |
469 \draw node[anchor=south east] at (-2.4,0.1) {\small $d=1$}; | 466 \draw node[anchor=south east] at (-2.4,0.1) {\small $d=1$}; |
470 \end{scope} | 467 \end{scope} |
471 \end{tikzpicture} | 468 \end{tikzpicture} |
472 \end{flushright} | 469 \end{flushright} |
473 | 470 |
474 \item \textsf{Catalogue Feature Vectors} - a sequence of track feature vectors for each track in the target instance ($catfeatures$) | 471 \item \textsf{Catalogue Feature Vectors} - a sequence of track feature vectors for each track in the instance ($features$) |
475 | 472 |
476 \item \textsf{Catalogue Unit Vectors} - a sequence of track unit vectors for each track in the target instance ($catunits$) | 473 \item \textsf{Catalogue Unit Vectors} - a sequence of track unit vectors for each track in the instance ($unitfeatures$) |
477 | 474 |
478 \item \textsf{Instance} - a catalogue, segmenter, feature, extractor, unit feature, unit extractor, dimension, catalogue feature vectors, | 475 \item \textsf{Catalogue Index} - a sequence of continuous interval indexes (one for each track). |
476 | |
477 An example can be seen below | |
478 | |
479 \[ \langle | |
480 \langle (0,r_{11}), (r_{11},r_{12}), (r_{12}, r_{13}) \rangle, | |
481 \langle (0,r_{21}), (r_{21},r_{22}), (r_{22}, r_{23}, (r_{23}, r_{24}) \rangle, | |
482 \langle (0,r_{31}), (r_{31},r_{32}), (r_{32}, r_{33}) \rangle, | |
483 \rangle \] | |
484 | |
485 \item \textsf{Instance} - a catalogue, segmenter, feature, extractor, unit feature, unit extractor, dimension, feature vectors, catal | |
479 | 486 |
480 % \emph{In our usage (see \textsf{Absolute} \and \textsf{Relative} in section \ref{s:refining}), we require certain semantics of the \textsf{Unit Feature} (and so of the values returned by its \textit{Unit Extractors}. Specifically, the values returned must be interpretable on a logarithmic scale in Bels, with the maximum possible value being the reference threshold at 0B.} | 487 % \emph{In our usage (see \textsf{Absolute} \and \textsf{Relative} in section \ref{s:refining}), we require certain semantics of the \textsf{Unit Feature} (and so of the values returned by its \textit{Unit Extractors}. Specifically, the values returned must be interpretable on a logarithmic scale in Bels, with the maximum possible value being the reference threshold at 0B.} |
481 | 488 |
482 \begin{schema}{Instance} | 489 \begin{schema}{Instance} |
483 cat : Catalogue \\ | 490 cat : Catalogue \\ |
485 seg : Segmenter \\ | 492 seg : Segmenter \\ |
486 f : Feature \\ | 493 f : Feature \\ |
487 x : Extractor \\ | 494 x : Extractor \\ |
488 uf : UnitFeature \\ | 495 uf : UnitFeature \\ |
489 ux : UnitExtractor \\ | 496 ux : UnitExtractor \\ |
490 d : Dimension \\ | 497 d : Dimension \\ |
491 catfeatures : \seq ~ ( \seq \V) \\ | 498 index : \seq ContIntIndex \\ |
492 catunits : \seq ~ (\seq \U) | 499 features : \seq ~ ( \seq \V) \\ |
500 unitfeatures : \seq ~ (\seq \U) \\ | |
493 \where | 501 \where |
494 d = f.fdimension \\ | 502 d = f.fdimension \\ |
495 catfeatures = map ~ (extract~seg~x) ~ cat \\ | 503 index = map~seg~cat \\ |
496 catunits = map ~ (extract~seg~ux) ~ cat \\ | 504 features = map ~ (extract~seg~x) ~ cat \\ |
505 unitfeatures = map ~ (extract~seg~ux) ~ cat \\ | |
497 \end{schema} | 506 \end{schema} |
498 | 507 |
499 \item \textsf{System Instances} - the set of System instances. | 508 \item \textsf{Singleton Instances} - the set of System instances which contain only one track |
500 | 509 |
501 Note that every instance must contain a catalogue in the music collection but not all catalogues in the collection will be part of an instance. | 510 \begin{schema}{SingletonInstance} |
511 Instance \\ | |
512 \where | |
513 \# cat = 1 | |
514 \end{schema} | |
515 | |
516 \item \textsf{System Instances} - the set of system instances. | |
517 | |
518 Note that every instance must contain a catalogue in the music collection but not all catalogues in the collection are necessarily part of an instance. | |
502 | 519 |
503 \begin{schema}{SystemInstances} | 520 \begin{schema}{SystemInstances} |
504 Collection \\ | 521 Collection \\ |
505 instances : \power Instance | 522 instances : \power Instance |
506 \where | 523 \where |
507 \{ i : instances @ i.cat \} \subseteq collection | 524 \{ i : instances @ i.cat \} \subseteq collection |
508 \end{schema} | 525 \end{schema} |
509 | 526 |
510 \section{Search} | 527 \section{Search Vectors} |
511 | |
512 \item \textsf{Search Vectors} | |
513 | |
514 Searching takes place using concatenations of feature vectors to build \emph{search} vectors. For a particular query all search vectors will have a fixed length equal to some multiple (which we will refer to as $sl$) of the dimension of the orignal feature vectors ($d$). | 528 Searching takes place using concatenations of feature vectors to build \emph{search} vectors. For a particular query all search vectors will have a fixed length equal to some multiple (which we will refer to as $sl$) of the dimension of the orignal feature vectors ($d$). |
515 | 529 |
516 In the schema below, the function $makesearchvs$ takes a natual number $sl$ and a sequence of feature vectors (associated with a track) to create a sequence of search vectors. The first element of the returned sequence is the concatenation of the first $sl$ feature vectors, the second element is also the concatenation the next $sl$ feature vectors but starting from the second element, and so on until all such sequences are formed. It should be clear that if the original sequence contains $n$ feature vectors then the output will contain $n-sl+1$ vectors. | 530 In the schema below, the function $makesearchvs$ takes a natual number $sl$ and a sequence of feature vectors (associated with a track) to create a sequence of search vectors. The first element of the returned sequence is the concatenation of the first $sl$ feature vectors, the second element is also the concatenation the next $sl$ feature vectors but starting from the second element, and so on until all such sequences are formed. It should be clear that if the original sequence contains $n$ feature vectors then the output will contain $n-sl+1$ vectors. |
517 | 531 |
518 \begin{flushright} | 532 \begin{flushright} |
611 \t1 sl > \# xs \implies makesearchvs~sl~xs = \langle \rangle \land \\ | 625 \t1 sl > \# xs \implies makesearchvs~sl~xs = \langle \rangle \land \\ |
612 \t1 sl \leq \# xs \implies makesearchvs~sl~xs = \\ | 626 \t1 sl \leq \# xs \implies makesearchvs~sl~xs = \\ |
613 \t2 concat (\langle (0 \upto sl-1) \dres xs \rangle) \cat makesearchvs~sl (tail~xs) | 627 \t2 concat (\langle (0 \upto sl-1) \dres xs \rangle) \cat makesearchvs~sl (tail~xs) |
614 \end{axdef} | 628 \end{axdef} |
615 | 629 |
616 \subsection{Identifying Source and Target} | 630 |
617 | 631 \item \textsf{Instance Search Vectors} |
618 A search vector is made from a source (sequence) and used to match against a user-defined target (instance). | 632 |
619 | 633 For any natural number less than the length of feature vectors for each track we can calculate the search vectors for an given by applying the $makesearchvs$ function to each of the sequences of feature vectors. We call this natural number \emph{search length} ($sl$) and we can use $map$ to apply $makesearchvs ~ sl$ to each of these sequences. |
620 \item \textsf{Source} - a track identified by the user in order to define a search. | 634 |
621 | 635 \begin{schema}{SearchVectors} |
622 \item \textsf{Target} - the instance used to match the source against. | 636 i? : Instance \\ |
623 | 637 sl? : \nat \\ |
624 \begin{schema}{IdentifySourceTarget} | 638 searchvs! : \seq (\seq \Vdsl) \\ |
625 source? : Track \\ | 639 unitvs! : \seq (\seq \Vsl) \\ |
626 tgt? : Instance \\ | 640 \where |
627 SystemInstances | 641 searchvs! = map ~ (makesearchvs ~ sl?) ~ i?.features \\ |
628 \where | 642 unitvs! = map ~ (makesearchvs ~ sl?) ~ i?.features \\ |
629 source? \in tracks \\ | |
630 tgt? \in instances | |
631 \end{schema} | |
632 | |
633 In addition the user can define the part of the source track to be used in the query. This is known as a sequence. A user defines the sequence by specifying the start and end points of the sequence index. (Note that there is an underlying assumption that the segmenter of the target returns a continuous interval index when applied to the source.) | |
634 | |
635 \item \textsf{Track Part} - a continuous sub-section of a track. | |
636 | |
637 \begin{zed} | |
638 TrackPart == Track | |
639 \end{zed} | |
640 | |
641 \item \textsf{Sequence} - the track part required for the search. | |
642 | |
643 \item \textsf{Sequence Index} - a continuous interval index that defines the sequence. | |
644 | |
645 \item \textsf{Sequence Length} - the number of intervals in the sequence index. | |
646 | |
647 \begin{schema}{IdentifySequence} | |
648 IdentifySourceTarget \\ | |
649 start?, end? : \nat \\ | |
650 seqindex : ContinuousIntervalIndex \\ | |
651 sl : \nat \\ | |
652 sequence : TrackPart | |
653 \where | |
654 start? \in \dom (tgt?.seg (source?)) \\ | |
655 end? \in \dom (tgt?.seg (source?)) \\ | |
656 seqindex = (start? \upto end?) \dres (tgt?.seg~source?) \\ | |
657 sl = end? - start? \\ | |
658 \end{schema} | |
659 | |
660 \subsection{Source Search and Feature Vectors} | |
661 | |
662 \item The \textsf{Source Feature Vectors} are the set of feature vectors of the sequence. | |
663 | |
664 \item The \textsf{Source Unit Vectors} are the set of unit vectors of the sequence. | |
665 | |
666 \begin{schema}{SourceFeatureVectors} | |
667 IdentifySequence \\ | |
668 sourcefeatures : \seq \V \\ | |
669 sourceunits : \seq \U \\ | |
670 \where | |
671 sourcefeatures = extract ~tgt?.seg~tgt?.x~sequence \\ | |
672 sourceunits = extract ~ tgt?.seg~tgt?.ux~sequence \\ | |
673 \end{schema} | |
674 | |
675 \item \textsf{Source Search Vector} - the concatenation of the feature vectors of the source sequence. | |
676 | |
677 \item \textsf{Source Unit Search Vector} - the concatenation of the unit vectors of the source sequence. | |
678 | |
679 \begin{schema}{SourceSearchVectors} | |
680 SourceFeatureVectors \\ | |
681 sourcesearch : \Vdsl \\ | |
682 sourcesearchunits : \Vsl \\ | |
683 \where | |
684 sourcesearch = (head~(makesearchvs~sl~sourcefeatures)) \\ | |
685 sourcesearchunits = (head~(makesearchvs~sl~sourcefeatures)) \\ | |
686 \end{schema} | |
687 | |
688 \subsection{Target Search Vectors} | |
689 | |
690 \item \textsf{Target Search Vectors} | |
691 | |
692 We can define these in exactly the same way as for the source, but as we now have a sequence of sequences of search vectors we have to use the map function to apply the $makesearchveors$ to each list | |
693 | |
694 \begin{schema}{TargetSearchVectors} | |
695 IdentifySourceTarget \\ | |
696 IdentifySequence\\ | |
697 targetsearchvs : \seq (\seq \Vdsl) \\ | |
698 targetunitsearchvs : \seq (\seq \Vsl) \\ | |
699 \where | |
700 targetsearchvs = \\ | |
701 \t1 map ~ (makesearchvs ~ sl) ~ tgt?.catfeatures \\ | |
702 targetunitsearchvs = \\ | |
703 \t1 map ~ (makesearchvs ~ sl) ~ tgt?.catfeatures \\ | |
704 \end{schema} | 643 \end{schema} |
705 | 644 |
706 \item \textsf{Hopped Search Vectors} | 645 \item \textsf{Hopped Search Vectors} |
707 | 646 |
708 Rather then generating all possible search vectors we may wish to general vectors starting at equally distanced intervals in the interval index of the tracks in an instance. Re refer to the size of this interval as $hop$. | 647 Rather then generating all possible search vectors we may wish to general vectors starting at equally distanced intervals in the interval index of the tracks in an instance. Re refer to the size of this interval as $hop$. |
709 | |
710 | 648 |
711 \begin{axdef} | 649 \begin{axdef} |
712 makehopsearchvs : \nat \fun \nat \fun \seq \Vd \fun \seq \Vdsl | 650 makehopsearchvs : \nat \fun \nat \fun \seq \Vd \fun \seq \Vdsl |
713 \where | 651 \where |
714 \forall xs : \seq \FV ; sl : \nat ; hop : \nat @ \\ | 652 \forall xs : \seq \FV ; sl : \nat ; hop : \nat @ \\ |
716 \t1 sl \leq \# xs \implies makehopsearchvs~sl~hop~xs = \\ | 654 \t1 sl \leq \# xs \implies makehopsearchvs~sl~hop~xs = \\ |
717 \t2 concat (\langle (0 \upto sl-1) \dres xs \rangle) \cat \\ | 655 \t2 concat (\langle (0 \upto sl-1) \dres xs \rangle) \cat \\ |
718 \t2 makehopsearchvs~sl~hop~ ((0 \upto (hop -1)) \ndres xs) | 656 \t2 makehopsearchvs~sl~hop~ ((0 \upto (hop -1)) \ndres xs) |
719 \end{axdef} | 657 \end{axdef} |
720 | 658 |
721 This enables us to define a hop size when generating target search vectors | 659 This enables us to define a hop size when generating search vectors |
722 | 660 |
723 \begin{schema}{TargetSearchHopVectors} | 661 \begin{schema}{HopSearchVectors} |
662 SearchVectors \\ | |
724 hop? : \nat \\ | 663 hop? : \nat \\ |
725 TargetSearchVectors \\ | 664 hopsearchvs! : \seq (\seq \Vdsl) \\ |
726 targethopsearchvs : \seq (\seq \Vdsl) \\ | 665 hopunitvs! : \seq (\seq \Vsl) \\ |
727 targethopunitsearchvs : \seq (\seq \Vsl) \\ | 666 \where |
728 \where | 667 hopsearchvs! = map ~ (makehopsearchvs ~i?.d ~ hop?) ~ i?.features \\ |
729 targethopsearchvs = \\ | 668 hopunitvs! = map ~ (makehopsearchvs ~ i?.d ~ hop?) ~ i?.features \\ |
730 \t1 map ~ (makehopsearchvs ~sl ~ hop?) ~ tgt?.catfeatures \\ | 669 \end{schema} |
731 targethopunitsearchvs = \\ | 670 |
732 \t1 map ~ (makehopsearchvs ~ sl ~ hop?) ~ tgt?.catfeatures \\ | 671 \item \textsf{Search Vector Power} |
733 \end{schema} | 672 |
673 For each Search Vector in an instances search vectors we take the associated unit values and take the arithmetic mean. | |
674 | |
675 \begin{axdef} | |
676 average : \V \fun \R | |
677 \end{axdef} | |
678 | |
679 \begin{schema}{Powers} | |
680 SearchVectors \\ | |
681 powers : \seq (\seq \R) | |
682 \where | |
683 powers = map~(map ~ average) ~ unitvs! \\ | |
684 \end{schema} | |
685 | |
686 \item \textsf{Search Vector Duration} | |
687 | |
688 The duration of the sequence of a the original feature vectors from which the search vector is composed. | |
689 | |
690 \begin{enumerate} | |
691 | |
692 \item \textsf{Single Search Vector Duration} | |
693 | |
694 This is defined as the duration of the sequence for each search vector. So for example if a search vector was made from the combination of 3 feature vectors and those 3 feature vectors came from the following intervals $ (3,8), (8,11), (11,13) $ then the duration of the search vector is $13-3=10$. | |
695 | |
696 \item \textsf{Track Search Vector Durations} | |
697 | |
698 Next we define a function at the level of track search vectors. We define a function which takes a sequence of search vectors for that track, the search length that was used to create them, and the track interval index and returns a sequence of natural numbers each which is the duration of the corresponding search vector | |
699 | |
700 \begin{axdef} | |
701 durationsTsv : \seq \Vdsl \fun ContIntIndex \fun \nat \fun \seq \R | |
702 \where | |
703 \forall svs : \seq \Vdsl; cii : ContIntIndex; sl : \nat @ \\ | |
704 \t1 durationsTsv~svs~cii~sl = \\ | |
705 \t2 \{ n : \nat | n \in 1 \upto n - sl + 1 @ \\ | |
706 \t3 (n, first (cii (n + sl)) - first (cii (n)) ) \} | |
707 \end{axdef} | |
708 | |
709 \item \textsf{Instance Search Vector Durations} | |
710 | |
711 Finally, we define a function called svdurations which takes a sequence of search vectors and calculate the track search vector durations for each sequence. | |
712 | |
713 \begin{axdef} | |
714 durationsIsv : \seq (\seq \Vdsl) \fun (\seq ContIntIndex) \fun \nat \fun \seq (\seq \R) \\ | |
715 \where | |
716 \forall svss : \seq (\seq \Vdsl); sl : \nat; ciis : \seq ContIntIndex @ \\ | |
717 \t1 durationsIsv ~ svss ~ ciis ~ sl = \\ | |
718 \t2 \langle durationsTsv ~ (head ~ svss) ~ (head ~ ciis) ~ sl \rangle \cat \\ | |
719 \t3 durationsIsv ~ (tail ~ svss) ~ (tail ~ ciis) ~ sl \land \\ | |
720 \t1 durationsIsv ~ \langle \rangle~ \langle \rangle ~ sl = \langle \rangle | |
721 \end{axdef} | |
734 | 722 |
735 \end{enumerate} | 723 \end{enumerate} |
736 | 724 |
737 | 725 \begin{schema}{Durations} |
738 \section{Distance, Power and Duration} | 726 SearchVectors \\ |
727 svdurations : \seq (\seq \R) \\ | |
728 \where | |
729 svdurations = durationsIsv ~ searchvs! ~ i?.index ~ sl? | |
730 \end{schema} | |
731 | |
732 % \end{enumerate} | |
733 | |
734 \section{Making a Query} | |
735 | |
736 \item \textsf{General Instance Query} | |
737 | |
738 Two instances, a source and a target are identified by the user for comparison as well as the search length from which the search vectors are defined for both the source and the sequence. Once the query has been defined we can generate the \emph{source search vectors}, \emph{source unit vectors}, \emph{target search vectors} and \emph{target unit vectors} as specified earlier. We also generate \emph{source powers}, \emph{target powers}, \emph{source search vector durations} and \emph{target search vector durations}. | |
739 | |
740 \emph{What are the pre-conditions regarding the search length? That it is less than the length of any sequence of feature vectors for every track?} | |
741 | |
742 \begin{schema}{GeneralQuery} | |
743 SystemInstances \\ | |
744 src?, tgt? : Instance \\ | |
745 sl?: \nat \\ | |
746 srcsearchvs!, tgtsearchvs! : \seq (\seq \Vdsl) \\ | |
747 srcunitvs!, tgtunitvs! : \seq (\seq \Vsl) \\ | |
748 sourcepowers, targetpowers : \seq (\seq \R) \\ | |
749 sourcedurations, targetdurations : \seq (\seq \R) \\ | |
750 \where | |
751 \{ src?, tgt? \} \subseteq instances \\ | |
752 \end{schema} | |
739 | 753 |
740 \subsection{Distance of Source Search Vector from Target Search Vector} | 754 \item \textsf{General Instance Self Query} |
755 | |
756 When an instance is identified as both the source and target we state that it is a self query. | |
757 | |
758 \begin{schema}{GeneralSelfQuery} | |
759 GeneralQuery | |
760 \where | |
761 src? = tgt? | |
762 \end{schema} | |
763 | |
764 \item Singleton (Track) Query | |
765 | |
766 A user defines an instance with exactly one track as the source. | |
767 | |
768 \begin{schema}{SingletonQuery} | |
769 GeneralQuery | |
770 \where | |
771 src? \in SingletonInstance | |
772 \end{schema} | |
773 | |
774 \item Point (Sequence) Query | |
775 | |
776 The user identifies a specific part of the track in a singleton instance to be used as the source known as a \emph{sequence}. A user defines the sequence by specifying the start and end points of the sequence index. Once these have been defined the sequence, sequence index and sequence length are easily identified. | |
777 | |
778 \begin{enumerate} | |
779 | |
780 \item \textsf{Source track} - the single track contained in the instance. | |
781 | |
782 \item \textsf{Sequence} - a continuous sub-section of the point query track used to make a query. | |
783 | |
784 \begin{zed} | |
785 Sequence == Track | |
786 \end{zed} | |
787 | |
788 \item \textsf{Sequence Index} - a continuous interval index that defines the sequence according to the segmenter of the instance. | |
789 | |
790 % \emph{Can we discuss this - is it reset so that the first element is indexed by 1 and so on?} | |
791 | |
792 \item \textsf{Sequence Length} - the number of intervals in the sequence index. | |
793 | |
794 \end{enumerate} | |
795 | |
796 \begin{schema}{PointQuery} | |
797 SingletonQuery \\ | |
798 start?, end? : \nat \\ | |
799 sourcetrack! : Track \\ | |
800 sequence! : Sequence \\ | |
801 seqindex! : ContIntIndex \\ | |
802 sl! : \nat \\ | |
803 \where | |
804 sourcetrack! = head (src?.cat) \\ | |
805 start? \in \dom (src?.seg (sourcetrack!)) \\ | |
806 end? \in \dom (src?.seg (sourcetrack!)) \\ | |
807 seqindex! = (start? \upto end?) \dres (src?.seg~sourcetrack!) \\ | |
808 sl! = end? - start? \\ | |
809 \end{schema} | |
810 | |
811 In this case the search vectors will be equal to a sequence containing only one sequence. That sequence will itself consist of only one sequence which will be a search vector of dimension $d * sl$ where $d$ is the dimension of the sequence and $sl$ the length of the sequence. | |
812 | |
813 \[ \langle \langle \Vdsl \rangle \rangle \] | |
814 | |
815 \section{Distance between Search Vectors} | |
741 | 816 |
742 We define the scalar product and distance between vectors of equal dimension. | 817 We define the scalar product and distance between vectors of equal dimension. |
743 | 818 |
744 \begin{axdef} | 819 \begin{axdef} |
745 scalarproduct : \V \fun \V \fun \R \\ | 820 scalarproduct : \V \fun \V \fun \R \\ |
748 | 823 |
749 \begin{zed} | 824 \begin{zed} |
750 \forall v_1, v_2 : \V @ distance~v_1~v_2 = 2 - scalarproduct~v_1~v_2 | 825 \forall v_1, v_2 : \V @ distance~v_1~v_2 = 2 - scalarproduct~v_1~v_2 |
751 \end{zed} | 826 \end{zed} |
752 | 827 |
753 Then we can define a variable which maintains a list of all distances between the source search vector and the target search vectors. | 828 Then we can define a variable which maintains a list of all distances between the source and target search vectors. If we have a sequence of sequence of search vectors in the both the source and the target we will end up with a sequence of sequence of sequence of sequences. |
754 | 829 |
755 \begin{schema}{Distances} | 830 For the purposes of illumination imagine a source and target each with two simple search vectors. Again, we will replace the reals with naturals here. |
756 distances : \seq (\seq \R) \\ | 831 |
757 SourceSearchVectors \\ | 832 Let us assume we have the following source search vectors for an instance of two tracks |
758 TargetSearchVectors \\ | 833 |
759 \where | 834 \[ \langle \langle a_1, a_2, a_3 \rangle, \langle b_1, b_2 \rangle \rangle \] |
760 distances = map (map (distance~sourcesearch)) targetsearchvs | 835 |
761 \end{schema} | 836 And the following target search vectors also for an instance of two tracks |
762 | 837 |
763 \subsection{Average `power' of a search feature vector} | 838 \[ \langle \langle x_1, x_2 \rangle, \langle y_1, y_2, y_3 \rangle \rangle \] |
764 | 839 |
765 For each Search Feature Vector in the target, we take the associated unit values and take the arithmetic mean. | 840 Then we need to generate the following data structure |
766 | 841 |
767 \begin{axdef} | 842 \[ \langle \\ % First track in source with target instance \\ |
768 average : \V \fun \R | 843 \t1 \langle \\ % First Search Vector with whole of target \\ |
844 \t2 \langle \\ % First Search Vector with first track | |
845 \t3 \langle a_1 \dot x_1, a_1 \dot x_2 \rangle, \\ | |
846 \t3 \langle a_1 \dot y_1, a_1 \dot y_2, a_1 \dot y_3 \rangle \\ | |
847 \t2 \rangle, \\ | |
848 % Second search vector | |
849 \t2 \langle \\ | |
850 \t3 \langle a_2 \dot x_1, a_2 \dot x_2 \rangle, \\ | |
851 \t3 \langle a_2 \dot y_1, a_2 \dot y_2, a_2 \dot y_3 \rangle \\ | |
852 \t2 \rangle, \\ | |
853 \t2 \langle \\ | |
854 \t3 \langle a_3 \dot x_1, a_3 \dot x_2 \rangle, \\ | |
855 \t3 \langle a_3 \dot y_1, a_3 \dot y_2, a_3 \dot y_3 \rangle \\ | |
856 \t2 \rangle \\ | |
857 \t1 \rangle, \\ | |
858 % Second track in source with target instance | |
859 \t1 \langle, \\ | |
860 \t2 \langle \\ | |
861 \t3 \langle b_1 \dot x_1, b_1 \dot x_2 \rangle, \\ | |
862 \t3 \langle b_1 \dot y_1, b_1 \dot y_2, b_1 \dot y_3 \rangle \\ | |
863 \t2 \rangle \\ | |
864 \t2 \langle \\ | |
865 \t3 \langle b_2 \dot x_1, b_2 \dot x_2 \rangle, \\ | |
866 \t3 \langle b_2 \dot y_1, b_2 \dot y_2, b_2 \dot y_3 \rangle \\ | |
867 \t2 \rangle \\ | |
868 \t1 \rangle \\ | |
869 \rangle \\ | |
870 \] | |
871 | |
872 | |
873 First let us define the function for a calculating the distances for a single search vector and general target search vectors | |
874 | |
875 \emph{I couldn't do this on one go first time round - too much nesting for me!} | |
876 | |
877 \begin{axdef} | |
878 vectordistances : \Vdsl \fun \seq (\seq \Vdsl) \fun \seq (\seq \R) | |
879 \where | |
880 \forall sv : \Vdsl; targetsv : \seq (\seq \Vdsl) @ \\ | |
881 \t1 vectordistances~sv~targetsv = map (map (distance~sv)) targetsv | |
769 \end{axdef} | 882 \end{axdef} |
770 | 883 |
771 \begin{schema}{Powers} | 884 Next we define the function for calculating the distances for the search vectors of a track with a target instance. |
772 targetpowers : \seq (\seq \R) \\ | 885 |
773 sourcepower : \R \\ | 886 \begin{axdef} |
774 SourceSearchVectors \\ | 887 trackdistances : (\seq \Vdsl) \fun \seq (\seq \Vdsl) \fun \seq (\seq (\seq \R)) |
775 TargetSearchVectors \\ | 888 \where |
776 \where | 889 \forall tracksv : \seq \Vdsl; targetsv : \seq (\seq \Vdsl) @ \\ |
777 targetpowers = map~(map ~ average) ~ targetsearchvs\\ | 890 \t1 trackdistances~tracksv~targetsv = \\ |
778 sourcepower = head (map~average ~ sourceunits) | 891 \t3 \langle vectordistances~(head ~ tracksv)~targetsv \rangle \cat \\ |
779 \end{schema} | 892 \t4 trackdistances~(tail ~ tracksv) ~ targetsv \land \\ |
780 | 893 \t1 trackdistances~\langle \rangle ~targetsv = \langle \langle \langle \rangle \rangle \rangle |
781 \subsection{Duration of a search feature vector} | 894 \end{axdef} |
782 | 895 |
783 Calculate the duration of the sequence for each search vector. (Made from the concatenation of feature vectors each associated with an interval. Add the intervals together to get the duration of the search vector.) | 896 Then we can define a function which calculating the distances between a source and target instance. |
784 | 897 |
785 \begin{schema}{Durations} | 898 \begin{axdef} |
786 targetdurations : \seq (\seq \R) \\ | 899 instancedistances : \seq (\seq \Vdsl) \fun \seq (\seq \Vdsl) \fun \seq (\seq (\seq (\seq \R))) |
787 sourceduration : \R \\ | 900 \where |
788 TargetSearchVectors \\ | 901 \forall sourcesv, targetsv: \seq (\seq \Vdsl) @ \\ |
789 SourceSearchVectors \\ | 902 \t1 instancedistances~sourcesv~targetsv = \\ |
790 \end{schema} | 903 \t3 \langle trackdistances~(head ~ sourcesv)~targetsv \rangle \cat \\ |
791 | 904 \t4 instancedistances~(tail ~ sourcesv) ~ targetsv \land \\ |
792 \subsection{Compiled Data} | 905 \t1 instancedistances~\langle \langle \rangle \rangle ~targetsv = \langle \langle \langle \langle \rangle \rangle \rangle \rangle |
793 | 906 \end{axdef} |
794 The we have a set of compiled data as follows. | 907 |
795 | 908 The next scheme takes a general query and calculates all the distances. |
796 \begin{schema}{CompiledData} | 909 |
797 Distances \\ | 910 \begin{schema}{CalculateDistances} |
798 Powers \\ | 911 GeneralQuery \\ |
799 Durations \\ | 912 distances : \seq (\seq (\seq (\seq \R))) |
913 \where | |
914 distances = instancedistances~srcsearchvs!~tgtsearchvs! | |
915 \end{schema} | |
916 | |
917 The list can be sorted into ascending order and some threshold set (as we shall see) where we believe it is sensible to expect some acoustic or cognitive similarity. | |
918 | |
919 Along with the distance between two search vectors, we also output the source track and index, the target track and index, as well as the duration the source search vector and power of the associated and corresponding unit vectors. | |
920 | |
921 \begin{schema}{Output} | |
922 srctrack, srcindex, tgttrack, tgtindex : \nat \\ | |
923 distance : \R \\ | |
924 sourceduration, targetduration : \R \\ | |
925 sourcepower, targetpower : \R \\ | |
926 \end{schema} | |
927 | |
928 The System outputs are a sequence of outputs ordered according to distance. We define a new variable which specifies the modified output that the user can make which we discuss in the next section. | |
929 | |
930 \begin{schema}{SystemOutput} | |
931 CalculateDistances \\ | |
932 output! : \seq Output \\ | |
933 modifiedoutput! : \seq Output | |
934 \where | |
935 \forall o_1, o_2 : Output @ \langle o_1, o_2 \rangle \inseq output! \\ | |
936 \t3 \implies o_1.distance \leq o_2.distance | |
800 \end{schema} | 937 \end{schema} |
801 | 938 |
802 \section{Refining a Search} | 939 \section{Refining a Search} |
803 \label{s:refining} | 940 \label{s:refining} |
804 | 941 |
805 System will first return the index of the track which gives the smallest distance from the source sequence search vector. Suppose this is at (3,4) in the list of lists. This signifies that the closest match is the 3rd track in the catalogue starting at the 4th interval. There are ways of refining the query. | 942 There are ways of refining the query. |
806 | 943 |
807 \begin{enumerate} | 944 \begin{enumerate} |
808 | 945 |
809 \item \textsf{Key List} - specify specific tracks within a catalogue to search over. | 946 \item \textsf{Key List} - specify specific tracks to search over. (Either source or target or both.) |
810 | 947 |
811 \item \textsf{Radius} - reject distances which are greater than a given real number radius. | 948 \item \textsf{Radius} - reject distances which are greater than a given real number radius. (Either source or target or both.) |
812 | 949 |
813 \item \textsf{Absolute} - reject any target search vectors where the `power' average is less than a specific absolute value. | 950 \item \textsf{Absolute} - reject any search vectors where the `power' average is less than a specific absolute value. (Either source or target or either or both.) |
814 | 951 |
815 \item \textsf{Relative} - reject any target search vectors where the `power' average is not within + or - a relative value. | 952 \item \textsf{Relative} - reject any search vectors where the `power' average is not within + or - a relative value. (Either source or target or either or both.) |
816 | 953 |
817 \item \textsf{Duration Ratio} - remove any search vectors whose total interval (duration) is sufficiently different from the duration of the source. | 954 \item \textsf{Duration Ratio} - remove any outputs where the relative durations of the search vectors are not within a specified range. |
818 | 955 |
819 \item \textsf{Hop Size} - Rather than making search vectors which start with the first feature vectors and then the second feature vector and so on, make sparser search vectors by starting with fv at 1, then fv at $1 + h$, then fv at $1 + 2h$ and so on where h is the hop size. | 956 \item \textsf{Hop Size} - Rather than making search vectors which start with the first feature vectors and then the second feature vector and so on, make sparser search vectors by starting with fv at 1, then fv at $1 + h$, then fv at $1 + 2h$ and so on where h is the hop size. (Either source or target or both with either equal or separate values of hop.) |
820 | 957 |
821 \end{enumerate} | 958 \end{enumerate} |
822 | 959 |
823 We specify what happens when values are given for any of these parameters. | |
824 | |
825 First, we specify a single output as signifiying a track, a starting interval, a distance, a power and a duration | |
826 | |
827 \begin{schema}{Output} | |
828 track : Track \\ | |
829 index : \nat \\ | |
830 distance : \R \\ | |
831 duration : \R \\ | |
832 power : \R \\ | |
833 \end{schema} | |
834 | |
835 The System outputs are a sequence of such outputs ordered according to distance. We define a new variable which specifies the modified output that the user can make. | |
836 | |
837 \begin{schema}{SystemOutput} | |
838 CompiledData \\ | |
839 output! : \seq Output \\ | |
840 modifiedoutput! : \seq Output | |
841 \where | |
842 \forall o_1, o_2 : Output @ \langle o_1, o_2 \rangle \inseq output! \\ | |
843 \t3 \implies o_1.distance \leq o_2.distance | |
844 \end{schema} | |
845 | |
846 \begin{enumerate} | 960 \begin{enumerate} |
847 | 961 |
848 \item Keylist | 962 \item Keylist |
849 | 963 |
850 \begin{schema}{KeyList} | 964 \begin{schema}{KeyListSource} |
851 SystemOutput \\ | 965 SystemOutput \\ |
852 k? : \power Track | 966 k? : \power \nat |
853 \where | 967 \where |
854 modifiedoutput! = output! \filter \{ o : Output | o.track \in k? \} | 968 modifiedoutput! = output! \filter \{ o : Output | o.srctrack \in k? \} |
855 \end{schema} | 969 \end{schema} |
856 | 970 |
857 \item Radius | 971 \item Radius |
858 | 972 |
859 The system removes any elements which have a distance greater than the radius. | 973 The system removes any elements which have a distance greater than the radius. |
864 \where | 978 \where |
865 modifiedoutput! = \\ | 979 modifiedoutput! = \\ |
866 \t1 output! \filter \{ o : Output | o.distance \leq r? \} | 980 \t1 output! \filter \{ o : Output | o.distance \leq r? \} |
867 \end{schema} | 981 \end{schema} |
868 | 982 |
869 \item Absolute | 983 \item Absolute Source |
870 | 984 |
871 \begin{schema}{Absolute} | 985 \begin{schema}{Absolute} |
872 SystemOutput \\ | 986 SystemOutput \\ |
873 a? : \R \\ | 987 a? : \R \\ |
874 \where | 988 \where |
875 modifiedoutput! = \\ | 989 modifiedoutput! = \\ |
876 \t1 output! \filter \{ o : Output | o.power \geq a? \} | 990 \t1 output! \filter \{ o : Output | o.sourcepower \geq a? \} |
877 \end{schema} | 991 \end{schema} |
878 | 992 |
879 \item Relative | 993 \item Relative |
880 | 994 |
881 %% \begin{axdef} | 995 %% \begin{axdef} |
882 %% abs : \R \fun \R | 996 %% abs : \R \fun \R |
883 %% \end{axdef} | 997 %% \end{axdef} |
884 | 998 |
885 | 999 |
1000 | |
886 \begin{schema}{Relative} | 1001 \begin{schema}{Relative} |
887 SystemOutput \\ | 1002 SystemOutput \\ |
888 rel? : \R \\ | 1003 rel? : \R \\ |
889 \where | 1004 % \where |
890 modifiedoutput! = \\ | 1005 % modifiedoutput! = \\ |
891 \t1 output! \filter \{ o : Output | abs (o.power - sourcepower) \leq rel? \} | 1006 % \t1 output! \filter \{ o : Output | abs (o.power - sourcepower) \leq rel? \} |
892 \end{schema} | 1007 \end{schema} |
893 | 1008 |
894 \item Duration Ratio | 1009 \item Duration Ratio |
895 | 1010 |
896 %%unchecked | 1011 %%unchecked |
897 \begin{schema}{DurationRatio} | 1012 \begin{schema}{DurationRatio} |
898 SystemOutput \\ | 1013 SystemOutput \\ |
899 d? : \R \\ | 1014 d? : \R \\ |
900 \where | 1015 \where |
901 modifiedoutput! = \\ | 1016 modifiedoutput! = \\ |
902 \t1 output! \filter \{ o : Output | \exp ^ {abs \ln (\frac{o.duration}{sourceduration})} \leq d? \} | 1017 \t1 output! \filter \{ o : Output | \exp ^ {abs \ln (\frac{o.duration}{srcduration})} \leq d? \} |
903 \end{schema} | 1018 \end{schema} |
904 | 1019 |
905 \item Hop Size | 1020 \item Hop Size |
906 | 1021 |
907 \begin{schema}{Hop} | 1022 \begin{schema}{Hop} |
908 hop? : \nat \\ | 1023 hop? : \nat \\ |
909 SystemOutput \\ | 1024 SystemOutput \\ |
910 TargetSearchHopVectors \\ | 1025 HopSearchVectors \\ |
911 \end{schema} | 1026 \end{schema} |
912 | 1027 |
913 \end{enumerate} | 1028 \end{enumerate} |
1029 \end{enumerate} | |
1030 \end{document} | |
914 | 1031 |
915 \section{Setting Thresholds} | 1032 \section{Setting Thresholds} |
916 | 1033 |
917 \subsection{For an Instance} | 1034 \subsection{For an Instance} |
918 | 1035 |
1016 | 1133 |
1017 \[ \langle V_{1}^{d}, V_{2}^{d} \dots V_{m}^{d} \rangle \] | 1134 \[ \langle V_{1}^{d}, V_{2}^{d} \dots V_{m}^{d} \rangle \] |
1018 | 1135 |
1019 When we apply this function to every track in a catalogue (which is a list of tracks) we simply determine a list of such lists. If there are $n$ tracks in a catalogue then our data would look something like the following. | 1136 When we apply this function to every track in a catalogue (which is a list of tracks) we simply determine a list of such lists. If there are $n$ tracks in a catalogue then our data would look something like the following. |
1020 | 1137 |
1021 \[ catfeatures = \langle ~~ \langle V_{11}^{d}, V_{12}^{d} \dots V_{1m_{1}}^{d} \rangle \\ | 1138 \[ features = \langle ~~ \langle V_{11}^{d}, V_{12}^{d} \dots V_{1m_{1}}^{d} \rangle \\ |
1022 \t3 ~~~ \langle V_{21}^{d}, V_{22}^{d} \dots V_{2m_{2}}^{d} \rangle \\ | 1139 \t3 ~~~ \langle V_{21}^{d}, V_{22}^{d} \dots V_{2m_{2}}^{d} \rangle \\ |
1023 \t3 \dots \\ | 1140 \t3 \dots \\ |
1024 \t3 \dots \\ | 1141 \t3 \dots \\ |
1025 \t3 ~~~ \langle V_{n1}^{d}, V_{n2}^{d} \dots V_{nm_{n}}^{d} \rangle | 1142 \t3 ~~~ \langle V_{n1}^{d}, V_{n2}^{d} \dots V_{nm_{n}}^{d} \rangle |
1026 ~~ \rangle | 1143 ~~ \rangle |
1027 ~~ \] | 1144 ~~ \] |
1028 | 1145 |
1029 The unit vector data has exactly the same structure but with different vectors all of dimension 1. | 1146 The unit vector data has exactly the same structure but with different vectors all of dimension 1. |
1030 | 1147 |
1031 \[ catunits = \langle ~~ \langle V_{11}^{1}, V_{12}^{1} \dots V_{1m_{1}}^{1} \rangle \\ | 1148 \[ unitfeatures = \langle ~~ \langle V_{11}^{1}, V_{12}^{1} \dots V_{1m_{1}}^{1} \rangle \\ |
1032 \t4 ~~ \langle V_{21}^{1}, V_{22}^{1} \dots V_{2m_{2}}^{1} \rangle \\ | 1149 \t4 ~~ \langle V_{21}^{1}, V_{22}^{1} \dots V_{2m_{2}}^{1} \rangle \\ |
1033 \t4 \dots \\ | 1150 \t4 \dots \\ |
1034 \t4 \dots \\ | 1151 \t4 \dots \\ |
1035 \t4 ~~~ \langle V_{n1}^{1}, V_{n2}^{1} \dots V_{nm_{n}}^{1} \rangle | 1152 \t4 ~~~ \langle V_{n1}^{1}, V_{n2}^{1} \dots V_{nm_{n}}^{1} \rangle |
1036 ~~ \rangle | 1153 ~~ \rangle |
1044 | 1161 |
1045 Also suppose that our sequence length $sl$ had a value of 2 and we applied the function $makesearchvs$. We would then make search vectors as follows. | 1162 Also suppose that our sequence length $sl$ had a value of 2 and we applied the function $makesearchvs$. We would then make search vectors as follows. |
1046 | 1163 |
1047 \[ \t3 \langle V_{1}^{d} \cat V_{2}^{d} , V_{2}^{d} \cat V_{3}^{d} \rangle \] | 1164 \[ \t3 \langle V_{1}^{d} \cat V_{2}^{d} , V_{2}^{d} \cat V_{3}^{d} \rangle \] |
1048 | 1165 |
1166 | |
1167 \end{enumerate} | |
1049 \end{document} | 1168 \end{document} |
1050 | 1169 |