Live data from Hacker News

Ditch Excel and Use Julia Data Frames

erik-engheim.medium.com

1–10 of 30 posts

Re: Ditch Excel and Use Julia Data Frames

#3
Spreadsheets automatically recalculate on partial updates to their inputs, which automates a lot of the coding that would be involved in data analysis. They are also a portable format that can be universally understood, which is much harder to say about dataframes in a typical business/work setting.

Re: Ditch Excel and Use Julia Data Frames

#5

Spreadsheets automatically recalculate on partial updates to their inputs, which automates a lot of the coding that would be involved in data analysis. They are also a portable format that can be universally understood, which is much harder to say about dataframes in a typical business/work setting.

The Pluto notebook for Julia does exactly the same. It creates dependencies between code cells and data automatically. If you change any data or code, that code ripples through through the rest of the notebook. So e.g. you could bind a slider to a variable and if that variable is used in a plot, dragging that slider will modify the plot in realtime.

Demonstration of Pluto.jl here:

https://www.youtube.com/watch?v=IAF8DjrQSSk

Re: Ditch Excel and Use Julia Data Frames

#6

People have been saying this for over 20 years now. Excel was, is and will remain a foundational aspect of compute tasks for programmers and non-programmers alike. It is not by chance that Excel is popular.

A lot of that is also simply momentum. Many people are utterly unaware of the alternatives outside the MS Office bubble. I have met so many people who thought the only program that exist for writing text is MS Word. Seriously. It is not even funny.

But that is how people often get trained early on. You get taught to use MS Word at school and other context and never get exposed to alternative tools or ways of doing things.

Here is a simple anecdote. Have on two occasions had long term Excel users get stuck on some task, looking for a solution. Then I ask them to use Apple Numbers. See if they can figure it out. Despite never having use it they figure out their task in no time.

My point is that MS Excel isn't even all that good of a spreadsheet application yet almost everybody use it despite better alternatives existing.

Re: Ditch Excel and Use Julia Data Frames

#7
IMO you're better off integrating excel into your workflow. You can use excel to just share the results as a report and view them as a report. Then when you want to do something with the data, just import it into your dataframe. Sometimes people are not ready to learn, sometimes organizations cannot be convinced to move forward because the cost of training an employee might end up more than the existing costs right now.

Re: Ditch Excel and Use Julia Data Frames

#8
What's with "ditch Excel"? Why not "ditch Pandas", which this is much closer to. Better yet, why not just create a tutorial on using Julia without mentioning other tools, and let the reader decide? Julia lets you SUM an entire column! Unbelievable! And you can rename a column? I'm blown away!

Excel has tons of point-and-click capabilities that don't require IMPORTing anything or knowing what an IOBuffer is. When you try to pretend that Julia can replace Excel, you just make yourself look foolish and uninformed.

Re: Ditch Excel and Use Julia Data Frames

#10
you can do this with pivot tables. if your database has more than 1m entries you can do some preprocessing with powerquery.

you don't even have to learn a single line of new code, if you've never coded before. you only need to know how to point and click with a mouse.

and the best thing is, you can take whatever you've done and send it down to whoever happens to have installed Excel on their machine, which, in a corporate environment, is everybody. Don't need to set up a new programming environment + dependencies.

Post reply on HN