I might be asking a question that has an opinionated answer but all opinions matter.
I was thinking on developing a cluster visualisation framework, but then I always wanted to know for a data scientist does visualisation matter more than number when it comes to cluster unless of course he/she wants to validate their algorithm. I have always wanted to visualise the data rather than looking at the number (could be Numerophobia, not). Is there any paper in psychology or data science that would give any reason for human perception more towards visually pleasing (the data) than plain old numbers?
Data visualization is better than numeric language because it gives a clearer picture of the results and need less effort for analysis in compared to numbers.
I think you want to compare tables vs other visualizations. However, tables are a type of visualizations. Although other visualizations (like charts, maps, photos,...) are mostly preferred by most of the people because they need less time to get the information, tables still have their importance especially:
- when the data are treated for individual values (like the results of an exam),
- when exact values are required (like threshold temperature of a process),
- when you need to compare values (like prices of some goods on a discount), etc.
- when you need to proof your trustfulness (as many other visualizations can be misleading, as shown in the attached picture (source: http://j.mp/2Ce8Ju3).
However, there are a lot of opinions that agree with you that other visualizations are better than tables (e.g. https://visual.ly/blog/avoid-tables-demand-visualizations/) ,
I conclusion, I highly encourage you to go on your project developing a cluster visualization framework, especially if it is for Python or R, and I would be happy to help in testing your work.
Best regards.
Refs:
- https://learnche.org/pid/data-visualization/tables-as-a-form-of-data-visualization
- http://www.storytellingwithdata.com/blog/2011/11/visual-battle-table-vs-graph
@james Ya that's true. Is there any literature review to that?
Please also check this link for additional info around your question.
https://www.researchgate.net/post/How_to_visualize_multivariate_data
“The greatest value of a picture is when it forces us to notice what we never expected to see” – John Tukey
Edward Tufte has written, designed, and self-published 4 classic books on data visualization.
You can find more details on his website https://www.edwardtufte.com/tufte/
Anscombe's quartet demonstrates both the importance of graphing data before analyzing it and the effect of outliers on statistical properties. Please refer to...
https://en.wikipedia.org/wiki/Anscombe%27s_quartet
- Badges
- Science topic