Preview only show first 10 pages with watermark. For full document please download

Binder 4

OLAP а П for : Online Analytical Processing 18 Т е г т coined in mid-1990's ã Main 9oal: support ad-hoc but complex querying performed Ь у business analysts CI Е е п г analysis to work with huge amounts of data in а data warehouse Definition: Set of processes and г that allow users to р е п о г т analytics across different entities and throughout different levels of detail using dlfferent dimensions

   EMBED

Share

Transcript

  з-о Cube Fact tabIe view: Multi-dimensional сиЬе: sa le prOdld storelddate р1 1 р2 ,1 1 р1 ,3 1 р2 с2 1 р1 ,1 2 р1 с2 2 .mt 12 11 506 .. 4 day2 day1 dimепsiопs = 3 J , ш Typical OLAP Operations up (drill-up, aggregation) :summarize/aggregate Clata D Ьу climbing uphierar chy of Ьу dimension redu CtQ n DгШ down (roll dow n): re\lerSE О' ГОII-uр ã from higher level sum mary to lower levelsummary 01 oeta , ed dati!, 01 Iп t гodLюп g newdimension s 51ice; д sUce is iI su~ of mult ,· d,me~lOna$ 3 <1., сor'esроnФng [ О iI $lnglE vallJf' 10 ' ()('If' 01 more member..ofthoedjlТler'\Sion~ r I ot;nthesub!.et Dice :The di(e operation is а slice оп rтю г е thantwodimensions of а datacube (ог moге than two consecutiveslices). Pivot (гo tate): ã reorient the cube, V1 sualization, зо to serie$ 01 20 planes Othef opeгabOГ\S drill ac ros s: i пvоl vi ng (across) more tt'lanone (ас! tabIedrill through : tflrough the bottom lellel of the cube { о its Ьдсk-ef1С1 relational tabIes,Some OlАР systeffiSalso allow yOU to jump ьас  totne sо uю ? data,You сап select data you findinteresting in thecubeand drill tnrough tothe source data to view extra detail 22 20 21 1!! ! Ш~[ ' Getting Multidimensional Data Out of DW : W ' ~ arge data vOlumes, e.g.,sales, telephone ca lls ã Giga-, Тега-, Peta-, Exa-byte ã OLAP = Оп-Liпе Ana lyt ical Processingã Interactive analysis dimепsiопs Example, findthe total sa l es (оуег ti me ) of each produd in each market SELECТ МаrkеСю, Product _ ю, SUM (Amount) FROM Sales ãExplorativediscovery ãFast г esponse times гequired ã OlAPoperationsã Output:ãAggregation of dataã Standard aggгegations operator,€.g ., SUM Mat1-;e l ID Pr cduct_ D SUM ~ Amount) ã Starting level,(Quarter,City) М1 Р1 500 ã Roll Up: less detail, Quarter - >Y eaг М1Р2 200 ã Dгill Down: Мorе detail, Quarteг -> Month М2 Р1 ~=·o ãSlice/Dice: Selectiоп, Уеаг= 1999ãDrill Acгoss: Join 2з 4