R was around in 2010, lol
Science Memes
Welcome to c/science_memes @ Mander.xyz!
A place for majestic STEMLORD peacocking, as well as memes about the realities of working in a lab.
Rules
- Don't throw mud. Behave like an intellectual and remember the human.
- Keep it rooted (on topic).
- No spam.
- Infographics welcome, get schooled.
This is a science community. We use the Dawkins definition of meme.
Research Committee
Other Mander Communities
Science and Research
Biology and Life Sciences
- [email protected]
- [email protected]
- [email protected]
- [email protected]
- [email protected]
- [email protected]
- [email protected]
- [email protected]
- [email protected]
- [email protected]
- [email protected]
- [email protected]
- [email protected]
- [email protected]
- [email protected]
- [email protected]
- [email protected]
- [email protected]
- [email protected]
- [email protected]
- [email protected]
- [email protected]
- [email protected]
- [email protected]
- !reptiles and [email protected]
Physical Sciences
- [email protected]
- [email protected]
- [email protected]
- [email protected]
- [email protected]
- [email protected]
- [email protected]
- [email protected]
- [email protected]
Humanities and Social Sciences
Practical and Applied Sciences
- !exercise-and [email protected]
- [email protected]
- !self [email protected]
- [email protected]
- [email protected]
- [email protected]
Memes
Miscellaneous
I learned how to do a fucking LOT of statistical shit in my degree. I also learned to get REALLY good at all kinds of shit in Excel.
Guess which helped my career on an actual practical way the most? Guess which made people seek me out at work for help with things?
Sometimes Excel is what's available. Sometimes it's just faster to do it that way rather than code up some ridiculously overdone solution in some programming language. Having both skills is best, but don't shit on opening an excel and just fucking getting it done, whatever it is.
If used right, it can also be a great equalizer with those less technically skilled in your workplace. You can quickly format and tune things and even layer a little bit of vba to make their lives easier without having to get into the complexity of an entire bespoke coded solution.
Also, a reminder for those in the back. For most of us, we aren't in college to learn a specific skill so much as we are there to learn how to be taught. To prove we are capable of taking instructions and producing results as requested.
If you never understand this, then you'll never understand later why you fail to land a high quality job.
we aren't in college to learn a specific skill so much as we are there to learn how to be taught.
I really like this idea, but prefer one small change: I think it's best to learn how to learn.
Learning how to be taught is part of that, and a large part. Understanding when to absorb information, rely on experts, and apply yourself until you improve is fundamental. You won't get any arguments from me there.
But being taught is only one facet of learning. Sometimes experts aren't really experts, or don't have the learner's best interests at heart, or omit things to protect their own interests or ideology.
Learning how to learn involves fostering fundamental curiosity, not being afraid to fail, asking all the questions even dumb ones or those with seemingly obvious answers. Finding out "why" something works instead of just "how". Fundamentally curious people who learn as a habit tend to also develop a scientific method-like approach to evaluating incoming information: "Ok, this is the information I'm presented with, let's assume the opposite, can I prove the null hypothesis?" This acts as a pretty good bullshit detector, or at the very least trains learners to be skeptical, to trust but verify, which is enormously important in the age of misinformation.
Being taught generally tapers off as someone gets older, or becomes an expert. Learning never needs to taper off, so long as your brain still works.
"Sometimes Excel is what's available."
I worked for a Big Company that was cutting back and dropped their Oracle contracts, forcing all the DBAs to work in Access. Then they fired all the DBAs, forcing everyone to either try to figure out Access or switch to Excel. Guess which way they went.
In my last job at that company, my department had built an Excel spreadsheet (database) so large and full of calculations they had to request money to update our machines to 64-bit Windows and 16GB RAM just to run it.
For most of us, we aren’t in college to learn a specific skill so much as we are there to learn how to be taught. To prove we are capable of taking instructions and producing results as requested.
This is true to the extent that you won't be solving Organic Chemistry 1 or Linear Algebra exercises at your workplace, but I think it's misleading. If anything, from my experience, people focus too much on producing the results and not enough on learning the skills. A lot of people stay on the mindset of "I only need the degree / where am I going to need that / the industry has moved on from this" and don't build strong foundations
FOSS is always available. R is always available. Your points remain but you're never in a situation where Excel is the only thing you can use.
Where I worked, many of the contacts specifically said we could not use open source software, so no, it is not always available.
You are if company policy dictates that's all you can use.
R, the language where dependency resolution is built upon thoughts and prayers.
Say what you want about Excel, but compatibility is kinda decent (ignoring locales and DNA sequences). Meanwhile, good luck replicating your R installation on another machine.
You heard about conda/containers/pixi/whatever?
PS: Excel will often fail if your system has a different default language. Like in many European countries one and a half is 1,5, not 1.5. Excel can't take it.
I have fucked around enough with R’s package management. Makes Python look like a god damn dream. Containers around it is just polishing a turd. Still have nightmares from building containers with R in automated pipelines, ending up at like 8 GB per container.
Also, good luck getting reproducible container builds.
Regarding locales - yes, I mentioned that. Thats’s a shitty design decision if I ever saw one. But within a locale, most Excel documents from last century and onwards should work reasonably well. (Well, normal Excel files. Macros and VB really shouldn’t work…). And it works on normal office machines, and you can email the files, and you can give it to your boss. And your boss can actually do something with it.
I also think Excel should be replaced by something. But not R.
This is more of an argument for LibreOffice (and in line with the post you're replying to) than it's an argument for using a programming language, let alone a specific one.
Current version of Excel stripped out a macro I made to make a specific task easier. It didn’t just block it from running. It refused to let me even see it anymore.
LibreOffice allowed me to see it again so I could re-implement it temporarily. I love how often Microsoft’s own tools can’t do what FOSS can with Microsoft files.
Lemmy Silver™🥈 incoming !
Excel I agree with.
But sometimes there is value in teaching the old tools/frameworks for doing something. For instance, in bioinformatics, I prefer students that can explain what the FASTA format is versus just boinking the pretty GUI button on the proprietary format used by their sequencer.
Not sure if the post is about GUI vs non-GUI. I read it as use R or pandas instead if SPSS.
I got bad news for this guy, his employer is only going to pay for excel and his coworker only knows how to use excel so he better learn to use excel. Also people do a lot of things in excel that have no business being done in excel.
I heard its good for databases...
I used to maintain an excel database along with an ecosystem of internal engineering tools in excel/vba. I worked in a vault, and one day I asked my isso if I could get python on some of the machines in my lab. A full 1.5 years later they got back to me that some security office was finally ready to consider my request and sent me a bunch of paperwork to fill out to justify why I needed python. And separate copies for each individual library I wanted to come with it. Needless to say I went on continuing to maintain my excel database and toolkit
some businesses just deserve to die, and are actively working towards it
A lot of companies are like this.
I'd hazard it's most companies.
My high school IT teacher said this outright. He was a FOSS guy, but he said employers will expect MS Office, so we're going to be learning that.
Funnily enough proprietary software is frowned upon in my professional domain. Im not mad though, the excel commands and whatnot still work in libre office spreadsheets.
Unrelatedly, doing statistics in a spreadsheet program sounds like absolute hell.
I have coded VBA-based shadow-IT back in the day, but I seriously think this is something that needs to disappear in most firms that still have it. It is typically unmaintable automation of tasks at the department level that is super dependent on who is around and is still often in use after the programmers are long gone. I have seen a few old VBA tools in use that should be done with standard python/R or god forbid even JS Code in a decent documented repo.
Should we also allow them to let AI write their essays?
If the math underneath is valid then I don't really care what calculator I use.
I'm no academic, but it seems wrong to me that any field would require the use of a particular proprietary software in order to do one's homework assignments.
May Excel or SPSS be the best tool for the job? In many cases, sure! But students should be allowed to use whatever other software can also get the job done, as long as the software exports the assignment in a data format that the professor can reasonably ingest (e.g.: turning in a CSV file, which can be understood by many different kinds of software, not just Excel).
I understand professors have limited time to check homework and thus don't want to spend time learning how to do anything but open a single, specific filetype, but that's besides the point.
Hmm. What about CAD? The professor going to teach FreeCAD, OpenSCAD, F360, OnShape, etc?
I think requiring one tool is OK. You're there to learn the process in a way that you can migrate to what you want later. Teachers aren't paid enough as it is, so it should be made as easy as possible for them to manage the flow of work.
Yeah, FreeCAD is open source and great (I use it at home)
But you're not landing a job requiring it, you'll use SolidWorks like everybody else
And if you find yourself in the 10% of companies which use some other CAD program, you will have to learn on the job
It is extremely common for classes to require students to learn to use proprietary software. It's a tool of the trade. If they graduate you without teaching you how to use it, they'd be fucking you over and ruining their own reputation. Like, imagine an accounting student graduating not knowing excel, because they did all their assignments using MatLab because they liked it better. It would be absolutely unthinkable for any potential employer to hire such a student. Excel is the software they use in that field. If a student wants to learn a different option in their free time, that's fine and dandy.
It is extremely common for classes to require students to learn to use proprietary software. It’s a tool of the trade.
I understand that; my position is more ideological than practical. In an ideal scenario, AutoDesk, Adobe, Microsoft, etc wouldn't be so deeply entrenched in their respective fields such that they are the de-facto tools of the trade for every business which must be learned in order to be hired. I know a given student has to learn certain proprietary tools in the current academic and professional environment. My comment was saying I would prefer this not to be the case. I am fully aware that proprietary software domination in the academic and professional spaces is not going away any time soon.
In my ideal scenario, an interviewer at a company would ask, "Can you perform the following edits to a given graphic?" instead of "Can you use Photoshop?" since the former allows for candidates who can use alternatives like GIMP. I understand company pipelines aren't set up for this, either, because company pipelines are also deeply entrenched in proprietary software.
The OP's photo is specifically about professors allowing other software to be used. Which would be a good starting point for making these kinds of changes.
What does not knowing excel even mean? If you're in STEM and would be tought the fundamentals of statistics independently of any software tool, you would gain the knowledge to apply these skills to any tool. And sorry to say, but after that Excel is just learning syntax.
True, but what I see in chemical engineering is college graduates not being very able to use Excel, because they all used Aspen and other very expensive software in college.
I can never say this enough: the best tool is the tool that gets the job done
Found the guy who would use a flathead screwdriver to regulate a demon core.
That reminds me of a story my bachelor's supervisor in astrophysics told me: One of his best PhDs applied at an insurance company. They got an Excel sheet with data that they had 1 week to analyze. All the other applicants took the whole week. He just put it in Python, solved it in a few hours, and got the job.
Old people and technology man. My advisor during my masters was an absolutely brilliant woman; she's one of the people who has been basically defining the field of data science since the early 90s. The first time I ever published with her, I sent my first draft and her response was "can you convert this to docx? I don't know how to work with tex." I still think she's one of the most brilliant people I've ever known but damn did it hurt to work on Microsoft word documents with her
R is best
Excel dominates the world of number crunching. Want to change that? You can't.
Imagine proposing an alternative for your organization that has been using Excel for two decades. Will every single sheet and workbook translate perfectly? If not, not good enough for dealing with numbers. This is why MS never fucks around with Excel. The risk/reward calculation (heh) is not a fit for open source spreadsheets.
Haha, Microsoft absolutely changes shit all the time. Millions of organisations around the world have an octogenarian computer standing somewhere in a corner because they need it running the old software.
Not Excel they don't, not ever. Been working in the office for almost 30 years. Excel does not change except to add tidbits and it's fully backwards compatible. You don't have any experience in this, do you?
LinkedIn ass title