Data Preparation Overview скачать в хорошем качестве

Data Preparation Overview

The Pyramid Decision Intelligence Platform uniquely combines Data Prep, Data Science, and Business Analytics with AI guidance in a no-code, business-user-centric application to power data-driven decision-making. Data Prep, the focus of this video, is the first stage in building a data factory for decision-making. It comprises of: data flow tools for fixing and blending data and semantic modeling tools for designing analytical data models ‘Data Flow’ is a visual, drag-and-drop interface for users to process data when good quality data is unavailable or when existing data needs adjustment. In Data Flow, you can: Connect to any data source and query its data Choose which tables and columns to import Clean data with widgets like filtering, deduplication, splitting, and trimming, Transform data with widgets like time intelligence, geocoding, and unpivot, Inject calculated values using point-and-click interfaces or inline functions, Deploy data science tools to add machine learning using built-in libraries like WEKA, MLIB, and Tensor flow. Or write learn-and-predict scenarios with Python and R. And finally, write the enriched data to one of the many supported data technologies, including Pyramid’s own in-memory database option. ‘Semantic’ or ‘Data modeling’ allows users to explain how their data is structured. These models are virtual and do not involve data ingestion. Instead, they are built against existing SQL data sources or databases created in a Data Flow. In Data Modeling: You can create complex table joins or let the built-in AI tools automatically detect and add the interconnections. You can set metrics and their formats using basic aggregations like sum and average. Or advanced options like distinct count and last-child And you can add hierarchies, including ragged and parent-child structures Ultimately, the semantic models drive the ultra-fast PYRANA query engine – the basis for resolving users’ data visualizations and sophisticated analyses without cutting code. To solidify data operations, ‘Master Flows’ allows you to construct advanced workflows to orchestrate multiple activities. Master flows include: Conditional, looping, and iterative executions, Interactions with external APIs and other apps, as well as internal Pyramid apps, Messaging and notifications. Combined with scheduling, master flows, and the prep framework are used to automate and productionize the data factory. Solid decisions are built on solid data. Democratizing access to data preparation amps up this process. Pyramid. For what’s next in data.