search

adilyalcin / Keshif

Keshif - Data Made Explorable (Prototype)
https://keshif.me
BSD 3-Clause "New" or "Revised" License
457 stars 123 forks source link

Export results #67

Open adilyalcin opened 9 years ago

adilyalcin commented 9 years ago

Exporting information from keshif for reporting/further analysis can be among the goals under some use cases. 1) Are such goals important and on point? 2) If export is needed and useful, what can be the exported data? 3) In which format / how to export selected information?

dlaehnemann commented 8 years ago

It would be great to not only be able to export subsets of the data, but to export / capture the current state of either an individual panel / plot or of the whole browser session with all panels, plots and selections visible into SVG or PDF to use for publications. Or am I missing an obvious way to generally do such an export from any browser?

adilyalcin commented 8 years ago

Hi @dlaehnemann,

Keshif currently does not have a feature to export images of individual views or the complete interface automatically. This feature is not among its current goals.

However, I understand that it can be useful to get screenshots, for publications and quickly sharing results.

On Mac, I commonly use the screen-capture shortcuts (https://support.apple.com/en-us/HT201361) It makes it really easy to automatically copy/save part/all of the screen so I can attach it to emails and publications. Some information for latest Windows version is here: http://windows.microsoft.com/en-us/windows/take-screen-capture-print-screen . I also like taking GIF shots to quickly share some selections and comparisons. Licecap (http://www.cockos.com/licecap/) is really useful for that.

Keshif is aimed to be a dynamic data exploration platform, and screen captures may not tell the whole story. As a hint, the screen captures generally need to be annotated or described externally.

Also, I will be happy to hear more about your use cases of Keshif, or what kind of exploration you commonly perform with it. Feel free to contact me here anytime,

Adil

dlaehnemann commented 8 years ago

Hi @adilyalcin,

thanks for the swift response. I am only just starting to try out your tool and I really like what I'm seeing -- so first of all thanks for your efforts in tackling the hard problem of set visualizations! My use case is in the analysis of sets of genetic variants across a set of different samples and I will definitely get back to you with more feedback later on.

For this suggestion here, it just came to my mind yesterday and I figured I'd just document it somewhere before I might forget it. Whenever I produce graphics for publication, I usually do so in some vector graphic format, in order to keep them as scalable as possible. Also, I prefer a vector graphic format for changing little details -- like highlighting important aspects of a figure, adding a legend or inserting labels. And finally, I try to get them published under a permissive license, to enable reuse, which obivously is most useful if the figure is a vector graphic.

Screenshots, however, will always be in some bitmap format. I figured that there has to be some cross-platform or cross-browser way to extract SVGs from a webpage, but haven't found anything. The only useful looking tool I came across is for Chrome: SVG Crowbar (https://nytimes.github.io/svg-crowbar/).

Thanks already, and I'll be back with further info on my use case(s)! :)

cheers, david