I asked this same question a couple of years back. I wonder: has the software that people use for visualization and data graphics changed at all? Punch your answer in the poll below. If you select ‘other’ let us know your tool of choice in the comments.
P.S. I know many of you use a combination of these. Pick your favorite if that’s the case.
{democracy:14}
Visualize This: The FlowingData Guide to Design, Visualization, and Statistics (2nd Edition)
SPSS
Constellation Roamer & Framework from Asterisq
I have used SPSS in the past, but I clicked excel. To be frank, I mostly use you!
Some of us use more than one of the languages/ programs listed. It doesn’t let us vote for more than one.
I use all of the above. That’s why I said to pick your favorite :)
How can you use “all of the above” being “nothing” one of them?
I have magic powers.
Everything I do starts with the data in an Excel spreadsheet and some initial cleaning/sorting/filtering/calculations there, so I felt obliged to click that. But Tableau Public is my favorite for easy online visualization and R for statistics and for power and flexibility in producing graphics. The following website is also great for quick and easy harnessing of R’s ggplot2 package for common statistical graphics: http://www.yeroon.net/ggplot2/. You can cut and paste the code for later customization in R itself.
agree completely Naomi…
I use SPSS/SAS/R and then Excel/Illustrator/Processing depending on volumes and what I’m doing…
IDV Solution’s Visual Fusion
What about adding Stata as an option?
Stata doesn’t count.
And why is Excel even on there?
No MATLABers?
MATLABer for life, although I’m under the impression I need to familiarize myself with R.
QlikView.
Great poll Nathan. Cheers.
Python obviously.
Indeed! with Matplotlib
STATA mostly
Another vote for python, usually accompanied by JavaScript and canvas or svg.
Stata
ArcGIS
JavaScript + SVG / Canvas, bien sûr!
Matlab
Spotfire. Their new free Silver offering publishes to the cloud:
http://spotfire.tibco.com/silverspotfire/default.aspx
Stata/R here.
Excel is not an analytical tool, poor reproducibility, poor statistical routines and poor graphics (adding a third dimension is totally unnecessary unless you have three variables you are wishing to graph).
Tons of reasons why, some links to discussions on mailing lists and peer-reviewed articles at http://slack.ser.man.ac.uk/progs/stata/avoid_excel.html
But you can make graphs with it, and I think that’s what most people do (even though there are better solutions).
Matlab
For me actually a mixture of Excel, Processing and in some rare cases R. But in most cases Excel is only for pre-analyzing and a first impression of the dataset, and for some exporting/importing stuff. Processing for the actual visual analysis and final visualization.
I use JMP, which is made by SAS. The high school site licence makes it very affordable, and the learners get to work with a program that is professional level.
Another vote for MATLAB
That Excel is “winning” currently somehow saddens me.
Here’s a positive spin: Excel only has 27% right now, so people are making use of other tools much more.
I jumped into this data world almost a year ago and have been doing most of my work in excel for a couple of reasons:
1. It’s what is on my computer
2. Most of the reports I pull from for my work come in two flavors: pdf or excel
3. Excel, since its on most computers, makes it easy to send a file or let others review my work
4. It’s what I know and have been working with since college
5. Graphs are simple and with features like conditional formatting heatmaps are practically built-in.
That being said I, I started reading this blog because I do want to get deeper into this field. So what would you recommend for someone who is comfortable with excel, has limited resources but wants to do more?
This list of tools and resources might be helpful:
https://flowingdata.com/2008/10/20/40-essential-tools-and-resources-to-visualize-data/
Matlab.
Photoshop, primarily for satellite images and maps. Otherwise I often use Excel or Aabel for graphing, then export a postscript file to Illustrator for cleanup.
Origin.
QlikView
@Rob Jensen: there is a free, fully functional personal edition available for download
I must say that Gephi.org is a must have for all graph visualizations and manipulation that happened to be my very first step when trying to explore a new dataset.
All,
Of interest to me is how the audience is organized in terms of profession and time on task doing data analysis and visual analytics vs or alongside data/info visualization. Lots of hammers here and lots of nails. What are folks actually building with the tools?
SAS neck and neck with pen/paper. Tableau whooping SAS?
Working on my own canvas-based lib ‘Unveil.js’
http://github.com/michael/unveil (work-in-progress).
Would enjoy some feedback/ :)
— Michael
Matlab!!!
I do a lot of data analysis with Ruby and Python. Otherwise it’s usually Excel, R and sometimes Processing.
I use Microsoft’s SSAS mainly these days, with either C#.net for weird stuff or Excel if it is light lifting. Most of my data sets are of the millions+ range of fact table rows.
Mathematica …
SAS is still my tool of choice for analytics.
@Rob Jensen,
We use a number of the tools in this list all day long, but more and more are finding that for commercial purposes, the Tableau (http://www.tableausoftware.com) suite of products cannot be beat. There is also a free version (Tableau Public – http://www.tableaupublic.com) which requires that you store your work in Tableau’s public cloud. Speed, flexibility (http://www.tableausoftware.com/learning/examples).
Full disclosure: We are Tableau evangelists of the first order, so take that for what it is worth. Check out the #Tableau and #TCC2010 stream on Twitter for some feedback from a wider audience of users.
Also, I am happy to have any offline conversation on the matter of tools for the tasks at hand.
MANY BLESSINGS!
Peace and All Good!
Michael W Cristiani
Analyze in SPSS.
Play with in Excel.
Visualize with actionscript and mxml in Adobe Flex
I also analyze in SPSS and (sometimes) play with in Excel.
ArcGIS
Mathematica.
python/ numpy/ scipy/ matplotlib xlswt occasionally to export to Excel
I use Matlab.
I notice that SPSS is not on the list and only has one vote in the comments. Do folks not use SPSS? I ask because I teach an undergrad stats class for psychology majors and I like to give them some exposure to data analysis software (other than excel).
Hello, Nathan! Bime Desktop – 1 hours, Esankey – 2 hours, Tableau – 14 hours, Excel – 68 hours. Also Parallel Sets, HCE and Google Docs Spreadsheet.
I use Gephi for exploratory network analysis.
I mostly use Excel, but I really like TinkerPlots for disaggregating and looking for patterns. And I also really like the latest visualizations that Google Spreadsheets has, especially the Gapminder-like motion graphs.
Oh, and ManyEyes. Love ManyEyes.
Igor Pro for graphing and analysis – awesome program.
Excel for some analysis and for “storing” data.
Matlab for when nothing else works well.
Igor Pro
SPSS (now PASW)
I’m happy to see ArcGIS up there though it’s not my primary.
I’ve used Spotfire for years and now it looks like they’re offering a free version with publishing for 1 year.
http://spotfire.tibco.com/silverspotfire/default.aspx
gnuplot
ArcGIS and MapInfo
Now Survo, a while ago it was SAS.
Protovis (Javascript) after cleaning up data with Excel.
I like NodeXL + Excel for social network analysis:
http://nodexl.codeplex.com
http://www.connectedaction.net
http://www.smrfoundation.org
Regards,
Marc
Gnuplot – quick and fast for just looking at stuff, but also can churn out publication-quality graphs if necessary (with some tweaking).
Also sometimes R and ggplot2, depending on what I’m doing.
1010data (www.1010data.com)
Chalk me up as another who still defaults to Matlab. Heck, I’ve even been known to write what are essentially shell scripts in Matlab because it’s just that easy. Inefficient, of course. But it beats figuring out the idiosyncrasies between tsch and bash, or, heaven forbid, learning Perl.
I keep meaning to migrate to Processing, Protovis, and/or Matplotlib. Just need the right excuse…
Anyone for Google Earth, MapServer, Octave, GMT or NCL?
Sure, it would be hard to imagine a world without excel but when it comes to grasping spatial variation in a dataset InstantAtlas is superb.
Mondrian – blows your socks off!
http://rosuda.org/Mondrian/
KNIME http://www.knime.org
http://www.eaglemap.com for oil and gas pipeline data
Python for everything.
Pingback: Encuesta de Flowing Data: ¿Qué usa usted para analizar y/o visualizar datos « Optimización y Estadística (SOE SC)