I Agree, your introductory paragraph nails it!
With any decision point where you are investing effort for a future benefit, there is individual and group inertia to change the status quo, but the benefits can be big.
I wrote a Medium article late last year about my 2020 project doing a Python deep-dive by researching and charting covid-19. Frustration with my mapping led me to evaluate tools for a few recurring needs, and for Data Viz I selected Plotly!
As I was working on pandemic analysis and plotting apps I carved out time to evaluate the components of my Python 'workbench' . I was frustrated devoting bandwidth to learning the esoterics of outdated tools. I needed to raise my productivity as well as output quality.
I wrote up eval criteria for IDE, Database, ETL, NLP, and Chart/Data Viz tools for Python. I would develop the same app using different tools for a direct comparison. The Plotly suite (express, studio, figure_factory, dash, etc) allowed me to do the most in the least amount of time. In terms of the domain, I was pulling daily and weekly public data sets, wrangling and synching it, then calculating various rates and ratios from it and plotting it as geodata. Matplotlib, which I first learned as the standard for my machine learning course, was so arcane. I felt like it had momentum with a lot of people, particularly in Universities, but it hadn't kept up. For maps with geodata, I tried several approaches like Google Maps API, matplotlib, and Plotly. I got the most done in the shortest time with Plotly.
The evaluations I did were a grind that I did on my own time. The negative was I had to learn packages only to decide they weren't my top choice and dump them. The Positive was the exposure to a big cross-section of Py packages gave me the fastest, broadest perspective across the Python development 'ecosystem'. Yeah, I spent last year 'drinking from the firehose' with Python.
One last Plotly positive: they have layers of documentation that cover all parameter definitions and value ranges plus code examples. I configured my IDE (PyCharm Pro) to access multiple sources of Plotly doc so I had fast and deep config help from a mouse hover. I know hover help is common, but it often leaves you guessing on extended parameters and or range of expected values.
Time savings by getting the exact info you need on hover adds up over dozens of times a day. It's time I wasn't asking questions on StackOverflow or doing trial and error to get the desired result!
That's a positive for both #PyCharm and #Plotly.
Anyway, thanks for posting your code to github too, good article.
