Showing posts with label Analytics DB. Show all posts
Showing posts with label Analytics DB. Show all posts

Wednesday, June 5, 2013

Projections in Vertica

Projections... You probably have a good idea of what that means already. Who remembers Plato's cave from high school? It's basically a group of people locked in a cave, staring at a blank wall all the time. All they see on that wall, are shadows of objects in the real world, projections if you will. Plato argued that, for these prisoners, these projections are as close as it gets to reality. However, people who reason about reality, and not just absorb it, free themselves from the cave. And can perceive reality as it really is. Not just its projections. 
In a relational database, you typically have tables, containing your data and its relations. This is reality. If you want to see it from a particular angle, you can project your data into a view. A view might be a subset of columns of a table or a combination of some columns of one table, with some other columns of another table. These things exist in Vertica as well, and they are called projections. But it pushes this notion one step further. In Vertica, there are no tables, only projections. And a collection of projections can represent a table, or multiple tables.  
So Vertica's idea of a projection is really Plato's cave turned inside-out. There is no reality. Only a collection of projections from which we can create that reality if we need to. Sounds familiar? 

Monday, April 29, 2013

Time series analytics on Vertica


Gap Filling and Interpolation (GFI)

A Swiss-Army Knife for Time Series Analytics

Gap Filling and Interpolation (GFI) is a set of patent-pending time series analytics features in Vertica . In this post, through additional use cases, we will show that GFI can enable Vertica users in a wide range of industry sectors to achieve a diverse set of goals.

Rolling Average with Oracle or Vertica analytical functions.


This little example will demonstrate how to use Oracle's or Vertica's analytical functions to get the rolling average. First you have to create and load a table that contains each month's average temperature in Edinburgh in the years 1764-1820.

Large-Scale Processing in Netezza.


Transitioning from ETL to ELT

CIO: Why is that uber-powered [commodity RDBMS] system running out of steam? Didn’t we just upgrade?
MANAGER: Yes, but the upgrade didn’t take.
CIO: Didn’t take? Sounds like a doctor transplanting an organ. Do you mean the CPUs rejected it? (laughing)
MANAGER: (soberly) No, just the users. Still too slow.
CIO: That hardware plant cost us [X] million dollars and it had better get it done or I’ll dismantle it for parts. I might dismantle your prima-donna architects with it!