Note: this is not a coding task.
I've just take a simple pandas series, and plotted a bar chart using each of the following:
Plot.ly is the only one that produced a chart that one can say was truly dynamic. And even then, the dynamism amounted to showing some values when the bars are moused over. And although some of the others provide the ability to zoom in, pan etc. and the technology is overall impressive and aesthetically pleasing, here's my issue:
Charts generally live, get distributed in, and get displayed in, static media such as Word, PDF, and PowerPoint, in the vast majority of corporate and academic settings. That's not going to change any day soon. Yes there are cases where things like maps etc. lend themselves to the sort of D3-like capabilities of zooming in and seeing more granularity. But on the whole, I am not seeing the value of producing Scatter, Box, Line, Column and other charts in Python.
Indeed, a suggestion I made recently to take some data and plot some of the above with Python was shot down, because the requester (a non-Python person), "would not be able to make revisions to the charts in Excel if she wanted". A very valid argument.
As a Python developer, I want to be convinced that really taking the time and effort to dive into creating visualisations in Python, and mastering it, is going to be worth it professionally. The task is to provide me with some real examples (either from your own experience, or from industry) of where Python visualisations are actually preferred (not just "a pretty alternative", but actually *preferred*) over Excel.
Examples from small, quirky, adventurous start-ups are not really what I'm interested in; we all know that these organisations are more adventurous and like the cutting edge of things. Nor am I interested in the examples of specialised, complex math based usage à la Matlab. I'm looking for examples of large-scale adoption in industry, or at least adoption *trends*.
The deliverable is either a document, or links to some white papers, or some of your own portfolio, or whatever you think is best. You decide. Again, I'm not interested in seeing any actual code. How many is "some examples"? I'm not counting the amount of examples you provide. I'm interested in being convinced. Convince me.
In your cover letter, just give me some indication that you've read the job spec and you understand what I'm after. I'm not looking to mine free advice - don't worry - so you can keep it short.