|
(Anything you wanna learn, you got it) I'm the gooniest Dataviz programmer of my professional US cohort, and I want everyone who likes graphs & maps to outperform people who publish books about their "data journey" This is a spiritual sequel to my fruitless "Goons learn to code!" GBS threads For the record, I owe my career to the SA goons who shat on my ambitions & called D3 hipster bullshit in 2014. Hope you've found comfortable retirement homes, my old friends
|
#
?
May 21, 2022 11:03
|
|
|
|
| # ? Apr 20, 2024 03:28 |
|
|
Data viz is cool and fun. On the flip side, few things are as disheartening as taking a request and creating the perfect dashboard only to have no one utilize it.
|
|
#
?
May 21, 2022 20:54
|
|
|
I have always found datavis an interesting subject, do you have any resources you would suggest reading for a high level overview of concepts, best practices, etc?
|
|
#
?
May 24, 2022 12:44
|
|
|
Vier posted:I have always found datavis an interesting subject, do you have any resources you would suggest reading for a high level overview of concepts, best practices, etc? Awesome! Tamara Munzner is the most academically credible DV person I know of, and these are recordings of her overview class https://www.youtube.com/watch?v=1GhZisgc6DI After that it depends on your medium. I almost never use Adobe design software or a sketch notebook and instead program visualizations to run in JS or Python. This book is well regarded by artists/designers, and the two books that let me bootstrap with D3 & React were by Scott Murray and then Elijah Meeks
|
|
#
?
May 24, 2022 16:56
|
|
|
I just got handed a data science or whatever role. The company I work for decided to stage all their data in a lake thing across decades of oracle / sap / ad hoc crap into a big thing and wants me to take up the datavis via tableau for our segment. Its a fortune 100 so...its a lot of data I imagine. They are also having me use knime...not sure what to expect from that except it looks like node red for spreadsheets. I guess I'm in this now.
|
|
#
?
May 25, 2022 23:15
|
|
|
Basic Poster posted:I just got handed a data science or whatever role. The company I work for decided to stage all their data in a lake thing across decades of oracle / sap / ad hoc crap into a big thing and wants me to take up the datavis via tableau for our segment. Its a fortune 100 so...its a lot of data I imagine. Don't drown. 
|
|
#
?
May 26, 2022 08:14
|
|
|
Good luck! I've avoided Tableau for 8 years, outside of one Data Visualization Engineer interview where the hiring manager sprung a live viz building challenge on me. Not an analyst here, buddy, I don't know this poo poo. Give me a D3 and React request and I'll whip something up
|
|
#
?
May 26, 2022 08:31
|
|
|
Is D3.js still the gold standard for data viz? I'm a data analyst who works mainly on Tableau and Jupyter Notebooks, but don't have any programming experience beyond mostly Python and a tiny bit of R. I've been using Voila and Panel along with ipywidgets to be able to generate any sort of visualizations that I don't usually do in Tableau (though I hate Tableau so very much and want to eventually move away from it entirely. I always forget which stupid pill combo needs to go into which stupid row or column to create the visualization I want instead of whoops here you go you just made a 100 x 100 row of individually generated map boxes along a grid and crashed the app). I've also been looking into Streamlit which seems actually way better than the above method, but for on-prem work I have to ask a dev to be able to deploy/build it for me (I'd learn how to myself but nothing internally is documented so while I could learn docker, kubernetes and whatever I have no idea what the setup even is to google for it). Do you use Streamlit at all, and if so, do you just deploy it on a cloud based platform or do you have your own servers you just build out the apps to? For a data analyst, is d3.js abstracted enough sort of like pandas for Python that there's lots of libraries and functions out there you can just use and point to your data or is it more like traditional javascript where you're coding out every single linear regression line or something? Or is it more for app developers who are making fully functioning apps that have data viz components rather than something useful for a business client facing data analyst who's given a dump of data and asked to make them a one-off dashboard or viz? Finally, what are your thoughts on ELK stacks, more specifically Kibana itself for data viz? Seems to be super popular for semi-structured data but its hard for me to figure out if its really something a non software engineer could use to do ad-hoc data viz or if its again, a stack for building out fully functioning apps that are data heavy. Oysters Autobio fucked around with this message at 12:47 on Aug 3, 2022 |
|
#
?
Aug 3, 2022 12:42
|
|
|
Oysters Autobio posted:Is D3.js still the gold standard for data viz? I'm a data analyst who works mainly on Tableau and Jupyter Notebooks, but don't have any programming experience beyond mostly Python and a tiny bit of R. Honestly, I don't know what the standard is for young-ish people in dataviz. And I want to know! The data visualization society is a very grassroots and (frustratingly) unambitious group built around education and community etc. If you can't join the discord then DM me for elevated customer support I love teaching (and got the only award in my PhD cohort for TAing), and I'm jaded after working "under" top-dataviz-nerd Elijah Meeks for 18 months, so you are in a very interesting position to learn everything I know about d3 & dataviz purely out of my spite. I will king/queen-make you if possible lol I coded these art projects with zero industry expectations and never learned a god drat thing to improve them from my "superiors" in 8 years. Yeah, of course I learned enough to fix and improve them, but I wouldn't be grumbling on SA if I had that work ethic you know? https://www.dropbox.com/s/5x8znms52ujdvl2/echoblock_moondog_2014.MOV?dl=0 https://bl.ocks.org/tomswisher/raw/09898dca3a6ca4ebfa33?raw=true https://bl.ocks.org/tomswisher/raw/09898dca3a6ca4ebfa33/?raw=true https://bl.ocks.org/tomswisher/raw/727113d072fad7a71eae6852efaf257e/?raw=true
|
|
#
?
Aug 7, 2022 14:38
|
|
|
.
Analytic Engine fucked around with this message at 00:53 on Jan 15, 2024 |
|
#
?
Aug 7, 2022 14:46
|
|
|
Looking at the examples of work you posted, data visualization seems like it's what I thought, basically ways of representing data in a visual format that you see on big news websites and that type of thing? Is an infographic a form of datavisualization? Could you discuss what job market is like for data visualizers? What type of companies hire them, how common are the jobs, what is typical compensation like?
|
|
#
?
Aug 10, 2022 03:09
|
|
|
cheque_some posted:Looking at the examples of work you posted, data visualization seems like it's what I thought, basically ways of representing data in a visual format that you see on big news websites and that type of thing? Those are critical questions I'll address soon, but I'm more interested in what you want to make and why you think dataviz might be the right medium for it. Generally though I consider dataviz to be anything where "factual"/"true" data is mapped to a visual, which includes both static graphics and animated gamelike bullshit. To me, an infographic is a static dataviz with many handmade visual decisions. I could draw things on my graphs but I strongly prefer to do *everything* in code, because I'm a nerd wannabe game designer. Infographics are pretty much on the Art end of the Art-Engineering spectrum for charts/graphs, and most maps are what I call infographics. Basically, if you can generate the piece without talking to an artist or graphic designer then it's not an infographic in my eyes Everything I make is an unfun videogame, which is absolutely not the perspective of infographicists who manually detail things without a system. I will add hours of code to make numeric labels works well; they used a stencil Analytic Engine fucked around with this message at 03:37 on Aug 10, 2022 |
|
#
?
Aug 10, 2022 03:29
|
|
|
Analytic Engine posted:Honestly, I don't know what the standard is for young-ish people in dataviz. And I want to know! The data visualization society is a very grassroots and (frustratingly) unambitious group built around education and community etc. If you can't join the discord then DM me for elevated customer support Thank you for posting this! I have some questions, if you're still around to answer. Forgive me, I know nothing about music. In the first link, what is the difference in the red and black? What is the difference in the tint/saturation of each? The other links don't work at this point. What advice do you have for someone who is just beginning to explore data visualization? What are the best (preferably free) resources to start out?
|
|
#
?
Jan 14, 2024 23:20
|
|
|
|
| # ? Apr 20, 2024 03:28 |
|
|
Reminding self to come back to this!
|
|
#
?
Jan 15, 2024 00:23
|
|
