In honor of the small technical question, for which this forum promised a safe haven from the beginning: I'm having a hard time figuring out how to map different words on a corpus of 11 texts using voyeurtools. So I want to map out surréalisme, surréaliste and surréalité. I imagine two graphs would be useful: a) One that would add them all up as a lemma; and b) One that would simply overlap them. I have seen Stéfan do it, so I know it can be done, but my figure-out-stuff-on-your-own switch is not working today for some reason.
How do you overlap different words on the same graph using voyeurtools?
(5 posts) (2 voices)-
Posted 4 years ago Permalink
-
I figured it out: You must select the words in the "Words in the Entire Corpus" module: http://twitpic.com/5u5g56
This feature seems really useful in lieu to account for variations in spelling, among other things. I confess I'm still in the dark about the possibility of a lemmatizer for voyeurtools.
Posted 4 years ago Permalink -
Stéfan Sinclair
AdminBest AnswerMy moment of euphoria at seeing a question about Voyeur Tools on DHAnswers (what better form of validation for a project?:) was short-lived as I realized that the answer would be unsatisfying: it's currently not possible to fuse multiple terms (or used a stemmed or lemmatized form) into a single expression (to get aggregate values). However, this functionality is part of the roadmap for this coming year and I hope we can show it working before DH12 in Hamburg.
Posted 4 years ago Permalink -
Here's another question, then, or rather a bug. Sorry, I don't know of another forum where I could do this. Might as well make it public here in case somebody else can benefit from it.
Of all the modules, the only one that seems to have a problem with the interface is "Words in the Entire Corpus." The search function and the heart selection tool overlap, so you can't select heart. There is a workaround, which involves reducing the size of the window, then using the expand button that comes up. This workaround then prevents you from using search, for which you have to expand again. In this back and forth, you can eventually select the necessary words for comparison. I don't know if this is an issue for Mac, but it is definitely an issue for all browsers in Windows 7 and Ubuntu. IE seems the best one, but it breaks down quickly. Here's a pic from Chrome: http://twitpic.com/5uawb5. I've had this problem since I started using voyeurtools.
By the way, voyeurtools FOR THE WIN! You have no idea how many improbable conclusions I've been able to collect using the tool on my texts. Today I found an almost perfect match in the distribution between the word "liberté" and "surréalisme" in a corpus spanning 5 years and 11 issues of the journal Tropiques. Even if you account for errors in transcription, and/or relative frequency depending on different contributors, the match holds. Long live statistics!
Posted 4 years ago Permalink -
Sorry about the UI glitch in that's preventing you from using the favourites correctly (I suspect that's a fairly infrequently used feature and I'm very glad you're trying to use it). I've seen this on occasion on a Mac, but it sounds like it happens much more frequently in Windows and Linux. I'll investigate further. BTW, did you try building your own skin? Maybe having only two columns would give more space for the UI elements to appear (while we try to find a more lasting fix). If you have problems with the skin builder, let me know (maybe by Twitter) what corpus you're using and I'll see if I can help.
Posted 4 years ago Permalink
Reply
You must log in to post.
