高效的空間洞察：LANDMARK的DATABRICKS和CARTO之旅.pdf

《高效的空間洞察：LANDMARK的DATABRICKS和CARTO之旅.pdf》由會員分享，可在線閱讀，更多相關《高效的空間洞察：LANDMARK的DATABRICKS和CARTO之旅.pdf（37頁珍藏版）》請在三個皮匠報告上搜索。

1、2024 Databricks Inc.All rights reservedJavier de la Torre,Founder and Chief Strategy Officer,CARTOJavier de la Torre,Founder and Chief Strategy Officer,CARTOJames Bentley,Data Scientist,Landmark Information GroupJames Bentley,Data Scientist,Landmark Information Group1EFFICIENT SPATIAL INSIGHTS:EFFIC

2、IENT SPATIAL INSIGHTS:LANDMARKS JOURNEY WITH LANDMARKS JOURNEY WITH DATABRICKS&CARTODATABRICKS&CARTO2024 Databricks Inc.All rights reserved2024 Databricks Inc.All rights reserved3A Brief IntroductionA Brief Introduction2024 Databricks Inc.All rights reserved2024 Databricks Inc.All rights reserved4Ge

3、oGeospatial CatCatalogue2024 Databricks Inc.All rights reserved5How far is this from the coast?How far is this from the coast?2024 Databricks Inc.All rights reserved2024 Databricks Inc.All rights reserved62024 Databricks Inc.All rights reserved2024 Databricks Inc.All rights reservedWhy H3 indexing?W

4、hy H3 indexing?7Endless*Universal TraversalPerfect Tessellation3 Is A Magic Number72024 Databricks Inc.All rights reserved2024 Databricks Inc.All rights reserved82024 Databricks Inc.All rights reserved2024 Databricks Inc.All rights reservedh3grid_distanceacg_ids62166610827014900008625621666108273164

5、000786256216661083005250003862562166610830173700038625621666108324741000186256216661083360780001862562166610834348400078625,12591621666108353445000554236216661083541660003862562166610837435100058625621666108542353000108625621666108600516000886256216661086033340003862562166610861906300028625,48429202

6、4 Databricks Inc.All rights reserved2024 Databricks Inc.All rights reserved102024 Databricks Inc.All rights reserved11Benefit#1Benefit#1General indexes enable General indexes enable arbitrary searchesarbitrary searches2024 Databricks Inc.All rights reserved2024 Databricks Inc.All rights reserved1220

7、24 Databricks Inc.All rights reserved2024 Databricks Inc.All rights reserved13h3namethreshold max_temp608433606150324000Perth and Kinross2526.1608410710333980000Northern Ireland2534.4608432126634754000Leicestershire2738.5608432199514980000Warwickshire2733.3608430290519457000South Yorkshire2629.36084

8、31576191074000Dumfries2525.7608432268486115000Gwynedd2528.5608431632210198000Cumbria2535.0608432258738552000Herefordshire2628.8608431922523144000Wiltshire2728.4608430877956898000Kent2728.5608431604913668000North Yorkshire2529.0608431126108700000Essex2735.12024 Databricks Inc.All rights reserved14Ben

9、efit#2Benefit#2Reduced computational Reduced computational requirementsrequirements2024 Databricks Inc.All rights reserved2024 Databricks Inc.All rights reserved1510 days,no flexibility2.5 hours,full flexibility2024 Databricks Inc.All rights reserved2024 Databricks Inc.All rights reservedPylonsPylon

10、s16DevaluationoVisual ImpactoSafety ImpactoPlanning RestrictionsoEMFDistance matters,but not precisely2024 Databricks Inc.All rights reserved2024 Databricks Inc.All rights reserved17100m2024 Databricks Inc.All rights reserved2024 Databricks Inc.All rights reserved1861.24m111.97m100m2024 Databricks I

11、nc.All rights reserved2024 Databricks Inc.All rights reserved19h3namethresholdmax_tempis_corechip_wkb608433606150324000Perth and Kinross2526.1TRUE608410710333980000Northern Ireland2534.4TRUE608432126634754000Leicestershire2738.5TRUE608432199514980000Warwickshire2733.3TRUE608430877956898000Kent2728.5

12、FALSEAQMAAAABAAAABAA608431576191074000Dumfries2525.7TRUE608432268486115000Gwynedd2528.5TRUE608431632210198000Cumbria2535.0TRUE608432258738552000Herefordshire2628.8TRUE608431604913668000North Yorkshire2529.0FALSEAQMAAAABAAAABwA608431126108700000Essex2735.1TRUE2024 Databricks Inc.All rights reserved20

13、24 Databricks Inc.All rights reserved202024 Databricks Inc.All rights reserved21Benefit#3Benefit#3Unification of Unification of different datasetsdifferent datasets2024 Databricks Inc.All rights reserved2024 Databricks Inc.All rights reserved222024 Databricks Inc.All rights reserved2024 Databricks I

14、nc.All rights reservedNot Just HexagonsNot Just Hexagons23UPRNUnique Property Reference NumberAddress,EPC,Sale historyH3EPSG:4326Distances,Aggregations,Neighboring DataBNGBritish National Grid|EPSG:27700UK Native DatasetsOther ways to indexOther ways to index2024 Databricks Inc.All rights reserved20

15、24 Databricks Inc.All rights reserved24Address MatchingLatitude/LongitudeCell IDGrid RefEastings/NorthingsMosaic TessellationCell Boundaries+ChipsBounding Box+ChipsAddressUPRNBNGGeomH32024 Databricks Inc.All rights reserved2024 Databricks Inc.All rights reservedOther Datasets.Other Datasets.25Landfi

16、ll DistanceTree DistanceMotorway DistanceSubsidence RiskSolar PotentialHeatstress RiskFlood RiskRasterVectorValue ModellingRisk ModellingRasterVectorValue ModellingRisk ModellingLand Value2024 Databricks Inc.All rights reserved26Benefit#4Benefit#4Better visualisation for Better visualisation for imp

17、roved stakeholder improved stakeholder engagementengagement2024 Databricks Inc.All rights reserved2024 Databricks Inc.All rights reserved272024 Databricks Inc.All rights reserved2024 Databricks Inc.All rights reserved282024 Databricks Inc.All rights reserved2024 Databricks Inc.All rights reserved292

18、024 Databricks Inc.All rights reserved2024 Databricks Inc.All rights reserved30Modern spatial Modern spatial analyticsanalyticsbuilt for the built for the cloudcloudAnalyze,visualize,and develop with spatial data at limitless scale inFully cloud nativeVisualize billions of data pointsPublish&share y

19、our insightsExtends Databricks geo capabilitiesAccelerate App developmentData enrichment2024 Databricks Inc.All rights reserved2024 Databricks Inc.All rights reservedBroad Platform EcosystemBroad Platform Ecosystem31“Flexibility”to choose your own adventure for geospatial processing“Flexibility”to c

20、hoose your own adventure for geospatial processingH3 integrationAnalytics Toolboxpydeck-carto visualizationSpatial Features for enrichmentDatabricks Data Marketplace+Whats best Whats best for my use for my use case?case?Databricks Labs Databricks Labs MosaicMosaic2024 Databricks Inc.All rights reser

21、ved2024 Databricks Inc.All rights reservedGrid indexing systems are ideally suited for scaleHierarchical system offers flexibilityEasy and effective visualizationScalable Geospatial Analytics with H3Scalable Geospatial Analytics with H332Supported natively in DatabricksSupported natively in Databric

22、ksPortfolio Risk AnalysisTelecom Network PlanningClimate Risk ModelingRetail Site Selection.Portfolio Risk AnalysisTelecom Network PlanningClimate Risk ModelingRetail Site Selection.Supporting the Databricks/Spark Supporting the Databricks/Spark ecosystemecosystemFastest in Databricks with Spatial S

23、QL and Photon acceleration,and Fastest in Databricks with Spatial SQL and Photon acceleration,and available in generic Sparkavailable in generic SparkCARTO BuilderCARTO WorkflowsCARTO Analytics Toolbox(UDFs and stored procedures)Databricks Spatial SQL BETAPhoton accelerationSpark58165DB SQLAll-Purpo

24、se ComputeWorking on Standard for a Geo Working on Standard for a Geo Data LakeData LakeBringing VECTOR RASTER into the Data Cloud via PARQUET.Bringing VECTOR RASTER into the Data Cloud via PARQUET.GeoparquetGeoparquet 1.11.1Vector data Vector data encoded in WKB or GeoArrow.On its way to become OGC

25、 standard.RasQuetRasQuet(Parquet Raster)0.1(Parquet Raster)0.1RasterRaster datadata encoded in native arrays or ZarrSupport for projectionsDifferent encodings Multidimensional35Preparing for the Table formats of Preparing for the Table formats of the futurethe future1.1.Universal Accessibility:Unive

26、rsal Accessibility:make raster data easily accessible on modern data platforms2.2.Integrated Data Ecosystems:Integrated Data Ecosystems:raster to coexist seamlessly with non-geospatial data in the same platform and prepared for modern distributionIf you are modeling your Data Lake,this is the way to

27、 add spatial data to it.Dont make GEO a special thingDont make GEO a special thing36SELECTRASTER_VALUE(raster,(11,23)FROM FloodRiskRasquetWHERE date=2024-05-16AND variable=RiskScore372024 Databricks Inc.All rights reserved38Javier de la TorreJavier de la TorreFounder and Chief Strategy OfficerCARTOCARTOJames BentleyJames BentleyData ScientistLandmark Information GroupLandmark Information Groupjames.bentleylandmark.co.ukStop by to chat at Booth#MP3(Marketplace Area)Stop by to chat at Booth#MP3(Marketplace Area)