4 ms·
Julia and excel are fundamentally different. In Excel, the logic/analysis part is on the same level as the data. The data and operations done on it are a singl
by alphachloride 6y ago
Julia and excel are fundamentally different.
In Excel, the logic/analysis part is on the same level as the data. The data and operations done on it are a single file. A datum is in a cell, and a formula is in a cell. You can move formulae around like you can move data around. You are viewing and operating on the data/logic in the same context. When you open an excel document, what you see right away is the culmination of all analyses done on the data.
With julia, the data and logic are separate entities. You have to first import data[1]. The result dataframe is created by explicitly running cells/scripts on the import. When I view the dataframe, I cannot just point-click-edit it. Any edit I make is a separate line of code to run. Then the whole script is run again over the whole dataframe to get the new result. This is opposed to excel which creates dependencies on a cell-level and live-updates the spreadsheet with each edit, but without processing the whole spreadsheet.
Since logic is its own entity (script) in julia, it is easier to write long and more complex processes: things you cannot easily do in a single-line formula parallelized over cells in excel. Also, since the processing and editing are explicitly spelled out in a script, it is easier to keep track of changes and reason about logic. So they both have their advantages.
[1]: You can even do that in excel. Import csvs, databases etc.
- whichquestion 6y agoAre there any existing FOSS tools that let you do a similar thing to what Excel does in the data layer but with Julia (or another programming language)? i.e. I’m in my cell and want to type my Julia right in that cell and watch it live update when I hit enter, similar to excel
- socialdemocrat 6y agoThe closest may be Pluto.jl https://www.youtube.com/watch?v=IAF8DjrQSSk https://www.youtube.com/watch?v=IAF8DjrQSSk In Pluto plots, calculations get bound to data and when you change that data the effects immediately ripple through. Personally if I work with table data in Julia, I will use it with something like Table Tool https://apps.apple.com/us/app/table-tool/id1122008420?mt=12 https://apps.apple.com/us/app/table-tool/id1122008420?mt=12 It lets you view and edit a CSV file. Saving and loading CSV files from Julia is very quick anyway. You can use a couple of key strokes to bring back a previous command. You also got some newer ones such as Tad: https://www.tadviewer.com https://www.tadviewer.com
- socialdemocrat 6y agoYes, I think the best approach is to combine tools. But what I think is a big problem is that people stretch the practical application of tools like Excel too far. They insist on using Excel far beyond its practical application. Personally I never use spread sheets, other than some simple plotting. I often prefer using a simple CSV viewer instead. I find it more practical to use tools which work natively with CSV, as that is the format one will be using in Julia, Python, R or whatever programming tools you use.