-1

I know that a data cube is a transforming multi-dimensional data which keeps changing with time, whereas data wrangling definition says it's transforming the data making it more valuable.

Isn't a data cube more meaningful and valuable piece of denormalized data ? I haven't been able to find any example to clear the symmetry, they both sound same to me.. Please help!

nitinr708
  • 1,393
  • 2
  • 19
  • 29
  • 2
    How is 'the process of transforming and mapping data from one "raw" data form into another format' the same as "pre-calculated multi-dimensional data" to you? – Jeroen Mostert Jul 28 '17 at 13:08
  • Funny. I meant technologically they seem to do the same thing, transform the data to be more valuable. And I asked for help understand from humans. You guys said, what a bot would say! Please give an example if possible – nitinr708 Jul 28 '17 at 13:14
  • 1
    I have never heard of the term "data wrangling" before today and I wouldn't use it, because it's too broad and vague. The value of a cube is obvious -- and a cube is a specific thing you can find examples of if you look. The value of transforming data from a format you can't use to one you can use, in the most general sense, is also obvious, and if you like you can call that data wrangling. But what, exactly, those transformations look like depend entirely on your problem domain, so direct comparison with a cube is meaningless. – Jeroen Mostert Jul 28 '17 at 13:24
  • There are a number of wrangling tools like spotfire in the market which are trying to replace the old cube processing players like datastage. If cube is specific version of data whereas wrangling is more dynamic, why won't you use it ? – nitinr708 Jul 29 '17 at 07:45
  • 1
    You misunderstand -- I have no opinion on tools that claim to do data wrangling, I'm just averse to the *term*. Tell me that a tool will do "data wrangling" for me and I still know nothing; I have to dive into the particulars to find out what it will and will not do, exactly. For all I know Spotfire may be the most wonderfullest thing in the history of data processing ever. From a *marketing* perspective, it's a very useful term; it sounds proactive and has associations with wild frontiers where data is wrangled rather than livestock. "Data analysis" is a broad, neutral term for the field. – Jeroen Mostert Jul 29 '17 at 15:04
  • I do not understand why would somebody downvote a question without understanding the intent behind asking it. Both are technologies meant for achieving data quality, There is nothing wrong in the question I suppose! – nitinr708 Aug 02 '17 at 14:22

1 Answers1

0

I found this article which claims to bring a perspective to this question -

I did not find true example but after reading and speaking to data analysts following line calls it a closure for me -

Data Wrangling is applied by functional experts on data in question to clean it off of it's veracity

On the other hand

Data Cube Processing is when a data analyst does a projection on structured data to output a report with some KPIs (Key Performance Indicators)

One is a 'cleanup' whereas another is a 'projection'

Community
  • 1
  • 1
nitinr708
  • 1,393
  • 2
  • 19
  • 29