3 ms·
> I have deep respect for Excel. I do something around the same line as you describe (build better tools in R/Python for larger data problems) but I have a dee
by ekianjo 7y ago
> I have deep respect for Excel.
I do something around the same line as you describe (build better tools in R/Python for larger data problems) but I have a deep aversion to Excel. It's proprietary, has horrible standards (it does not support native UTF8 csv files for example), and is stuck in the 80's in terms of paradigm. And this is precisely the tool that prevents people from doing things more efficiently because "they can do it manually in Excel even if it takes a long time". It makes people take really, really bad habits.
- formercoder 7y agoIn what way is the paradigm stuck in the 80s?
- astine 7y agoWell if you're using Python, you're following an imperative/oop paradigm which goes back to the sixties, so I don't know why you'd criticize Excel for using a dataflow model. I can understand not wanting to use Excel because it's proprietary (but Libreoffice isn't) but I can can understand not wanting to deal with a lot of it's legacy cruft, but for the stuff that Excel does well, it does it a lot more quickly, more interactively, and more intuitively for most people than a similar solution in Python. In fact, one of the key benefits of an Excel solution is that you can see every step of your calculation which creates a degree of self documentation which you don't get with Python, for example.
- ekianjo 7y ago> that you can see every step of your calculation which creates a degree of self documentation which you don't get with Python, for example. If you want to do that in Python then Jupyter Lab/notebook is a good solution to document your code and see the result of each operation. And way faster than Excel for large data tables.
- BeetleB 7y agoYep. My first job had a lot of Excel. I hated it. Once I learned pandas, I rarely needed Excel again for my work.
- phonon 7y ago> it does not support native UTF8 csv files It does though? You can export as a UTF-8 csv file...check the export options.
- ekianjo 7y agoI am talking about importing CSV files in UTF8. If I use Japanese characters for example it fails utterly to display properly unless I import it as "Text" first.
- jchb 7y agoThe secret trick for importing data into Excel (without actually writing a file in the Excel file format): don't try CSV - as you have already noticed that will fail in all kind of fragile ways depending on your character encoding and operating system locale. Instead, convert your data into an html table and then paste or import that into Excel. But not any html table, you need to have the appropriate proprietary css properties in there, such as the mso-data-placement:same-cell. Those are actually documented ...in the Office HTML and XML Reference published in 1999.
- chaseha 7y agoI frequently go from SQL query results to HTML table to excel in a similar manner, but I haven't had issues w/ any proprietary CSS properties. Using Paste Special -> Match Destination Formatting it always comes out without any issues
- phonon 7y agoNo, you should use the Data import tools--that's what they're there for!
- phonon 7y agoYes, it takes a little longer, but you should use Data > From Text/CSV for more complex text/delimited imports.
- taneq 7y agoThink of all the open, modern, nicely standards compliant tools that you use. Now think of the fact that Excel lets people solve their problems more easily than all the things you like. I'm not saying it's good, but it's better than everything else (for the people who use it.)
- halfeatenpie 7y agoExcel is a tool. It empowers those who don't know how to code to do their jobs more effectively than before without "learning to code". It's something you keep in your garage in case you need it. It depends on what their current workflow is. Sometimes I'll use Excel, sometimes I'll use R/Python. What's important is that Excel gives you access to the computing power of modern technology with a lower barrier-of-entry/knowledge. That's one of the most valuable things that many people fail to understand. Not everyone is a programmer and for a lack of a better term, "doesn't really care about UTF-8 and just want the calculations to work". Also, in many cases Enterprise IT in critical infrastructures are resistant to supporting R or Python due to their security models (firewall rules prevents install.packages('tidyverse') for example). This is what makes us specialists and someone they bring on board to help them out and explain to them about what the "best-use-case" is, and then watch as people try to cram a square peg into a round hole.