http://forums.flowingdata.com/ Strength in Numbers en Sat, 11 Feb 2012 18:34:26 +0000 http://forums.flowingdata.com/topic/creating-graphs-in-illustrator-quick-question#post-2511 Tue, 24 Jan 2012 18:02:49 +0000 eelert 2511@http://forums.flowingdata.com/ <p>Hi!</p> <p>It does seem possible to change the shape of data points in Illustrator. Adobe has a nice tutorial: <a href="http://www.adobe.com/designcenter/illustrator/articles/illcs2at_chart_print.html" rel="nofollow">http://www.adobe.com/designcenter/illustrator/articles/illcs2at_chart_print.html</a> </p> http://forums.flowingdata.com/topic/visualizing-large-social-network-graph-data#post-2452 Tue, 27 Dec 2011 13:26:27 +0000 nathany 2452@http://forums.flowingdata.com/ <p>This thread might be helpful:</p> <p><a href="http://forums.flowingdata.com/topic/large-network-layout-and-visualization" rel="nofollow">http://forums.flowingdata.com/topic/large-network-layout-and-visualization</a> </p> http://forums.flowingdata.com/topic/visualizing-large-social-network-graph-data#post-2450 Mon, 26 Dec 2011 23:07:24 +0000 firesofmay 2450@http://forums.flowingdata.com/ <p>Hi,<br /> I am looking to create a large Social Networking Graph using the data I have scraped using Python from facebook.</p> <p>What I am interested in is, to Show the interconnections between people, and the nodes should be clickable, and when you hover over the nodes, it highlights the direct connections of them.</p> <p>The data set I am playing with is quite Large, Million Nodes give and Take.</p> <p>I have found static ways to generate the data, but I am looking for is more of interactive view of it.</p> <p>Can anyone suggest me the best to do this?</p> <p>Thanks a lot. I really appreciate any kinda help :) </p> http://forums.flowingdata.com/topic/phdpostdoc-genomic-visualization-in-belgium#post-2211 Thu, 26 May 2011 05:00:45 +0000 jandot 2211@http://forums.flowingdata.com/ <p><strong>Background</strong></p> <p>The last decade has seen a tremendous evolution in the field of clinical genetics. However, although new high-throughput DNA sequencing techniques have opened the door for routinely identifying genome variations that might contribute to a given disease, there are still several bottlenecks that stand in the way of widespread application in the clinic. Several of these issues can be (partly) solved with data visualization techniques.</p> <p>Geneticists rely heavily on genome browsers to investigate genetic variation in their genomic context (see e.g. <a href="http://www.ensembl.org" rel="nofollow">http://www.ensembl.org</a> and <a href="http://genome.ucsc.edu" rel="nofollow">http://genome.ucsc.edu</a>; visualization expert Ben Fry created <a href="http://bit.ly/lBK6wR)" rel="nofollow">http://bit.ly/lBK6wR)</a>. Although of priceless use, these browsers all have the significant drawback that they can only display data relative to a so-called "reference". It is also impossible for them to show rearranged chromosomes as they appear in for example cancer.</p> <p><strong>Aim</strong></p> <p>Within this project, we will build a <strong>new interactive genome browser</strong> that is capable of displaying so-called structural genomic variation and is able to display patient data relative to actual control (healthy) data rather than a theoretical reference.<br /> This visualization will include several features. First of all, it will be possible to visualize the differences in chromosome structure between individuals. This will be based on a directed graph, and might be inspired by the ABySS-Explorer tool by Nielsen et al (<a href="http://bit.ly/iZhLi1)" rel="nofollow">http://bit.ly/iZhLi1)</a>. Secondly, the tool will enable any path through this graph to be considered as the "reference", so that depictions resembling the genome browsers such as <a href="http://bit.ly/l6W0dV" rel="nofollow">http://bit.ly/l6W0dV</a> can be created. Thirdly, it will be possible to simultaneously look at high resolution while still maintaining the context of the whole chromosome. Finally, this tool will be able to visualize different levels of uncertainty in the data.</p> <p><strong>Profile</strong></p> <p>The ideal candidate for this position holds a MSc or PhD degree related to data visualization. He/she should be eager to learn about genomics and variation in the human genome. He/she will collaborate with PhD students and postdocs both at ESAT and the University Hospital in Leuven, as well as with clinical geneticists and international collaborators. The position requires good analytical skills and knowledge of programming languages such as java, ruby or perl. Knowledge of data visualization frameworks such as Processing (<a href="http://www.processing.org" rel="nofollow">http://www.processing.org</a>) and/or Protovis (<a href="http://vis.stanford.edu/protovis/" rel="nofollow">http://vis.stanford.edu/protovis/</a>) are a significant plus. The position should be filled as soon as possible.</p> <p>We offer a competitive package and a fun, dynamic environment connected to a top-notch consortium of young leading scientists in human genetics and cancer. The University of Leuven is one of Europe's leading research universities, with English as the working language for research. Leuven is one of Europe's most beautiful university towns, just outside Brussels, at the heart of Europe.</p> <p>To apply for this position, please send your CV together with a photograph and - if possible - some examples of your work to Prof Jan Aerts (jan.aerts@esat.kuleuven.be). The deadline for this application is 30th June 2011. </p> http://forums.flowingdata.com/topic/relative-probability-comparing-several-samples#post-1966 Sun, 24 Oct 2010 15:17:29 +0000 drbeechwood 1966@http://forums.flowingdata.com/ <p>Hi,<br /> First post. I use "relative probability diagrams" which are similar to weighted histograms that show the abundances of populations of ages of minerals within a certain rock.</p> <p>Data: I analyze 100 grains, that range in age from say 200 million years to 60 million years. there might be 80 grains that are 70 +/- 5 million, then another smaller group of ages at 100 +/- 5 million years. etc. Each age has it's own uncertainty, so the data are weighted based on abundance and uncertainty. The area under each curve is equal. </p> <p>The curves themselves are fine, but it is cumbersome to display and compare multiple samples because of the height of the curves. I thought about some kind of stacked area chart, but maybe someone else has a better idea.</p> <p>The first image shows the output from the program.</p> <p><a href='http://img23.imageshack.us/i/plot1.jpg/' rel="nofollow"><img src='http://img23.imageshack.us/img23/8061/plot1.jpg' /></a></p> <p>The second image is when I extracted the curves and put them on a common axis, I may have also reduced the vertical scale by 50%.</p> <p><a href='http://img176.imageshack.us/i/plot2.jpg/' rel="nofollow"><img src='http://img176.imageshack.us/img176/9194/plot2.jpg' /></a></p> <p>Anything else come to mind that would help easily compare the main peaks on these curves between samples?</p> <p>If images don't load I will post them in a reply, as there is no preview post option.<br /> Thanks! </p> http://forums.flowingdata.com/topic/social-network-mapping#post-1937 Sun, 10 Oct 2010 23:01:32 +0000 nathany 1937@http://forums.flowingdata.com/ <p>this could be a good place to start:</p> <p><a href="http://forums.flowingdata.com/topic/large-network-layout-and-visualization" rel="nofollow">http://forums.flowingdata.com/topic/large-network-layout-and-visualization</a> </p> http://forums.flowingdata.com/topic/social-network-mapping#post-1925 Tue, 05 Oct 2010 12:45:03 +0000 Henky 1925@http://forums.flowingdata.com/ <p>Hey everyone!</p> <p>My name is Henk. I'm from Holland and i've been searching the net forever to find a programm. But I have almost lost all hope of something like it exsisting!:) So I hope one of you can point me in the right direction!</p> <p>I want to create a social network graph/map. I don't have any programming skills whatsoever. Im looking for a program that lets you input names of friends and the weight of the relationship they have with one another.<br /> Than it would be awesome if I could update this data every month and see a evolving/moving graph or map of how my friends grow closer or apart from one another!</p> <p>Does any programm like this exist??</p> <p>Hope to hear from you all:D</p> <p>Greetz from Holland,</p> <p>Henk </p> http://forums.flowingdata.com/topic/freelance-node-edge-graph-project#post-1713 Thu, 06 May 2010 17:18:14 +0000 roloc 1713@http://forums.flowingdata.com/ <p>Looking for someone to build a node-edge interactive graph module using the Prefuse Flare, BirdEye RaVis, or similar actionscript libraries. The front-end will need to be driven by a simple database (even a google spreadsheet) that employees at my company will maintain by hand. </p> <p>If you're interested, email Eben at:</p> <p>egilfenbaum[at]ceip[dot]org </p> http://forums.flowingdata.com/topic/how-to-create-logarithmic-graph#post-1703 Thu, 29 Apr 2010 15:58:10 +0000 Novasphere 1703@http://forums.flowingdata.com/ <p>@ ivyleaf</p> <p>Actually, the logarithmic function in Excel is very easy, here's the formula:</p> <p>=LOG10(cell reference with your original data)</p> <p>It gets more complicated if the scale is not base 10, but that's Excel's default, so probably this will work. You may have to play with your divisions a bit, but using a screen shot of an Excel graph as your guide should help.</p> <p>Cheers! </p> http://forums.flowingdata.com/topic/how-to-create-logarithmic-graph#post-1682 Mon, 12 Apr 2010 09:49:23 +0000 ivyleaf 1682@http://forums.flowingdata.com/ <p>Thanks for both your help! </p> http://forums.flowingdata.com/topic/how-to-create-logarithmic-graph#post-1679 Thu, 08 Apr 2010 21:43:25 +0000 Bobmcconn 1679@http://forums.flowingdata.com/ <p>could you export a nice graph from Excel (or whatever) as a postscript or pdf file and pull that into Illustrator? It's cheating, in a way, but way easier.. </p> http://forums.flowingdata.com/topic/how-to-create-logarithmic-graph#post-1673 Wed, 07 Apr 2010 11:07:39 +0000 nathany 1673@http://forums.flowingdata.com/ <p>you'd have to convert the y-values to the log scale (in Excel perhaps) and then put it in to illustrator. </p> http://forums.flowingdata.com/topic/how-to-create-logarithmic-graph#post-1672 Tue, 06 Apr 2010 22:25:15 +0000 ivyleaf 1672@http://forums.flowingdata.com/ <p>Hi, I'm new to this forum. I just started a new job and have been tasked to figure out how to create a logarithmic graph in Illustrator... I can create a linear graph in Illustrator, but am at a loss on how to do a log-log or semi-log graph outside of Excel. Any help you can offer is greatly appreciated!</p> <p>Thank you! </p> http://forums.flowingdata.com/topic/advice-for-some-time-series-data-visualization#post-1438 Fri, 22 Jan 2010 13:27:41 +0000 nathany 1438@http://forums.flowingdata.com/ <p>I second that. </p> http://forums.flowingdata.com/topic/advice-for-some-time-series-data-visualization#post-1435 Fri, 22 Jan 2010 11:26:52 +0000 joshhemann 1435@http://forums.flowingdata.com/ <p>This might be obvious, but a heatmap works best when you impose some order to your variables. Time has a natural ordering but you might sort the 40 countries by GDP, those in NATO vs not, etc, and hopefully blocks of colors jump out and reveal information more easily. You should post your images when you are done! </p> http://forums.flowingdata.com/topic/advice-for-some-time-series-data-visualization#post-1434 Thu, 21 Jan 2010 17:55:15 +0000 nathany 1434@http://forums.flowingdata.com/ <p>I guess while we're at it, a treemap/mosaic plot or whatever it's called, could work too. </p> http://forums.flowingdata.com/topic/advice-for-some-time-series-data-visualization#post-1433 Thu, 21 Jan 2010 05:48:58 +0000 New High Score 1433@http://forums.flowingdata.com/ <p>You can also look at the concept of Marimekko charts, which are a similar concept. </p> http://forums.flowingdata.com/topic/advice-for-some-time-series-data-visualization#post-1423 Sat, 16 Jan 2010 01:06:09 +0000 kerimcan 1423@http://forums.flowingdata.com/ <p>Thanks for your help. Heatmaps are what I was looking for. I found especially the second link in krees' message useful. </p> http://forums.flowingdata.com/topic/advice-for-some-time-series-data-visualization#post-1422 Fri, 15 Jan 2010 10:29:52 +0000 nathany 1422@http://forums.flowingdata.com/ <p>i guess a regular time series chart might work for this too. </p> http://forums.flowingdata.com/topic/advice-for-some-time-series-data-visualization#post-1421 Fri, 15 Jan 2010 02:55:08 +0000 Kim Rees 1421@http://forums.flowingdata.com/ <p>Perhaps you're thinking of a basic heatmap?</p> <p>I don't use R, but one of these looks like it might help:<br /> <a href="http://sekhon.polisci.berkeley.edu/stats/html/heatmap.html" rel="nofollow">http://sekhon.polisci.berkeley.edu/stats/html/heatmap.html</a><br /> <a href="http://enotacoes.wordpress.com/2007/11/16/easy-guide-to-drawing-heat-maps-to-pdf-with-r-with-color-key/" rel="nofollow">http://enotacoes.wordpress.com/2007/11/16/easy-guide-to-drawing-heat-maps-to-pdf-with-r-with-color-key/</a> </p>