And its good because Excel is a horrible tool. Much better spend your time learning a language aimed at data analysis.
I don’t need c++ or Python to build a net present value analysis that I could show the CFO. I can build it in excel in an afternoon, and I can hand it to the least technical salesperson, and he can make changes in my model and see the results instantly then send it back to me. And then my boss can. And then someone in marketing can.
And the bonus is I don’t get a bunch of useless errors that no one understands like “Stack Overflow Error” or “Cannot process because of error 1994505 super duper data transformation canooter value fluctuation.” The worst error I get is that it crashes and I lose 10 mins of work.
Face it - Excel is a phenomenal achievement and it burns arrogant programmers like you because we can do all this and completely shut you out. That’s why it’s a “horrible tool” lol.
Already Lotus 1-2-3 for MS-DOS, more than 30 years ago, matched or exceeded almost all features provided today by MS Excel (and it had an optimized keyboard-based user interface that enabled experienced users to perform most tasks faster than in modern spreadsheet programs like Excel or LibreOffice Calc).
Excel has just provided the same features, with nothing original of any importance, but only with trivial improvements enabled later by better computers.
The fact that Excel has succeeded to replace Lotus 1-2-3 has not been caused by any technical superiority but by the integration in MS Office and by the fact that Microsoft has cheated, by ensuring that nobody else could keep up with the changes in the public Windows API and with the undocumented parts of the API.
Only someone who isn’t a heavy user of Excel could say this.. the improvements really aren’t trivial.
It is obvious that a program which may have a size of many megabytes is able to include a lot of improvements over a program whose size was limited to a few hundred kilobytes.
Nevertheless, all improvements provided by Excel are quantitative not qualitative, they do not enable any essentially new application.
For instance, there is no doubt that it is much easier to write a maintainable Excel script in Basic, than in the awkward macro language of Lotus 1-2-3.
Even so, using just the macro language of Lotus 1-2-3, it was possible to write amazingly complex applications, e.g. for the automation of the tracking in real time and of the generation of reports about the flow of partially processed products through the steps of complex technological processes in some factories (with thousands of different products and with more than one hundred manufacturing process steps through which they might have to pass, depending on the part number), also of the printing of all the documents that accompanied the manufacturing batches, and where the data was stored in a database embedded in the spreadsheets.
Today, writing such a program by using a modern database, a modern programming language and a modern computer would be very easy, but doing the same within 640 kB and with a 33 MHz CPU was not a little accomplishment. At that time, Lotus 1-2-3 was one of the most useful applications for most businesses, especially when computers and programs were much less affordable and many would not have been able to buy any other program, except perhaps some word processor.
Excel has taken all the features of Lotus 1-2-3, except the user interface, which was no longer suitable in Windows, but everything added later were just enhancements that were obvious when faster CPUs and more memory became available.
I'd be willing to bet you also probably look down on people who use Windows instead of Linux desktop. Or why they don't "burden" themselves to learn the oh so easy command line. If this is you, I get exactly the person you are and it defends my point even more.
And I didn’t mean that Excel is bad or anything, it’s extremely good at what it’s for. And like u mention, it’s much easier for someone to pick up excel and “get stuff done” than say python. My point was simply that there’s merits to one suggesting an alternative to excel related tasks. It’s not unequivocally a better choice for everyone, but it certainly can be.
And lol windows is more so just the privacy thing and msft fucking their users around. I’ve heard great things about powershell. But ironically, the command line rlly isn’t difficult to grok, and certainly not hard relative to working with ridiculously nested XLOOKUP’s
It's not just analysis; data representation, manipulation and collaboration are essential parts as well (especially the last one). Being able to quickly import numbers, format and surface the important part and show it to someone to discuss and tweak in real time makes a world of difference.
I’d bet money you are also the person who wonders why the world uses Windows over Linux for many tasks. When you can answer why regular people choose Excel over code and why regular people choose Windows over Linux, you’ll get it, and you can make a positive change in code to bring Linux and Code more mainstream.
Generally tho idk if it’s a good idea to optimize for usability from sales dude
in matplotlib? get outta here
The only thing that comes even close for my current Excel spreadsheets use cases would be Rshiny. Interactive responsive analysis of various scenarios, with traceable computation for a very wide audience, is far better via Excel than even Rshiny, which hides all the code.
Starting with Excel is probably a great way to get to other tools. Especially when computations are straightforward and don't require the advanced stats libraries available in something like R.