Visualizing Data

  • Comments posted to this topic are about the item Visualizing Data

  • There are several things I like about physical maps

    • The ability to see a area much larger than the one your prime interest lies.

  • The element of pre-planning
  • The skill in using a map
  • The sheer delight in discovering something interesting, following the trail and discovering something that wasn't marked on the map
  • GPS devices take a huge amount of that away from you.

    The question I have with data visualisation is whether they can take things too far?  Isn't one of the most rewarding things following a line of thought and discovering something new?  Would you be more convinced by a visualisation that gave you a high percentage of what you were looking for but, through your own efforts, you made the final discovery?

    If you spoon feed someone the final answer do they gain understanding and do they buy in to the answer?

    Perhaps it is a valuable skill to be capable of putting together a visualisation that leads towards a conclusion but maintains the polite fiction that the conclusion was anything other than inevitable?

  • I think there are different visualizations for different purposes.
    - we need information quickly to make a decision. a GPS tends to be better here in the moment. Turn left in 200 ft.
    - We need more information to be change decisions quickly. A GPS fails at times here, depending on the implementation. If I need to make a detour or there's a lot of traffic and I turn now, what are my options. In some sense a map is better at rerouting (With a passenger's help).
    - We want information about the larger picture.  For example, planning a trip. A GPS is not as good as a real map. The interface, resolution, scale in/out quickly are slower for me with digital maps. It's easier to work on a physical item and move my eyes than keep moving the screen. Maybe that's a generation thing?
    - We want to sell or convince someone of a story. Digital visualizations can help here. A GPS with street view might be better here.
    - We are doing research about possibilities. Again, I think either can work here, but it's easier to change digital visualizations.

  • David.Poole - Tuesday, March 21, 2017 2:22 AM

    There are several things I like about physical maps

    • The ability to see a area much larger than the one your prime interest lies.

  • The element of pre-planning
  • The skill in using a map
  • The sheer delight in discovering something interesting, following the trail and discovering something that wasn't marked on the map
  • GPS devices take a huge amount of that away from you.

    The question I have with data visualisation is whether they can take things too far?  Isn't one of the most rewarding things following a line of thought and discovering something new?  Would you be more convinced by a visualisation that gave you a high percentage of what you were looking for but, through your own efforts, you made the final discovery?

    If you spoon feed someone the final answer do they gain understanding and do they buy in to the answer?

    Perhaps it is a valuable skill to be capable of putting together a visualisation that leads towards a conclusion but maintains the polite fiction that the conclusion was anything other than inevitable?

    I have found that younger generations who have never used a physical map and have only used GPS devices do not seem to learn the roads as well. I don't mean the familiarity of how to drive them but the knowledge of the routes between places. They, generally, follow the SatNav without thought or deviation. To do so would be unthinkable.

    Because of that they tend to be at a disadvantage when their SatNav device fails as they have not learnt the skills of reading street signs for directions to the same level nor do they usually have the confidence to try a route.

    These are terrible generalisations but my experience validates this (even considering nearest and dearest so I am not being disparaging - at least not intentionally so). Basically, it is an example of the overreliance on technology by some.

    Gaz

    -- Stop your grinnin' and drop your linen...they're everywhere!!!

  • I think that the most interesting data visualisations to come will be those yet to be imagined (or from Sci-Fi as pointed out).

    I shudder at the thought of the future point in time when I pass a clothing store to turn and see a digital mannequin resembling me to try and sell me an outfit. :crying:

    Gaz

    -- Stop your grinnin' and drop your linen...they're everywhere!!!

  • Visualization for most of those working in either data analyzation or data science fields are not using the visualization as the final conclusion. It's just one piece of the bigger pie, but a important piece because it's easier to communicate with pictures than raw data.

    For example, visualization in our data science department (the one I work with exclusively), uses visualization to help tell a story. Some important considerations are used with that story that reflect: what the data is telling us, how you should feel, and most importantly, what recommended actions you should take. This all requires a story where insights are unearthed, methodologies are explained on how we got to these conclusions, and of course, visualizations to help communicate everything in between.

    The interesting thing though. Most of the enterprise tools out there like Tableau, MicroStrategy and even PowerBI have not touched the power of what the language based modules/libraries can do with R, Python and SAS. While they require much more scripting/programming, the customization you can achieve with Matplotlib, Seaborn and more is insanely good for creating immersive visualization pieces that drive the overall story home.

  • Viewing 6 posts - 1 through 5 (of 5 total)

    You must be logged in to reply to this topic. Login to reply