Closed jamiewaese closed 11 years ago
Yes. We can scale the whisker width, 100 being double the width of the chromosome. I can place it as a world object so that its size and coordinates get updated automatically with camera movements.
Double may be too small for the maximum value. Maybe triple or quad.
On Oct 4, 2013, at 11:22 PM, Hans Yu notifications@github.com wrote:
Yes. We can scale the whisker width, 100 being double the width of the chromosome. I can place it as a world object so that its size and coordinates get updated automatically with camera movements.
— Reply to this email directly or view it on GitHub.
Triple and quad can be 150 and 200. There won't be a maximum (or minimum) limit! Or, maybe the width should be 1, then double would be 2, triple would be 3, etc. That would make a bit more sense. Of course, decimals would be supported too.
I've already wrote a class for user annotations. Now I just need to confirm whether the annotations would be created using the IDENTIFIER:WIDTH:COLOR notation or something else (like the color choices on annotation popup). I plan to combine the annotation system with marking genes by the Get Data button, i.e. genes annotated will be loaded just like hitting the Get Data button for that gene. I really like the idea of having the color choices on the annotation popup, but how can we make it consistent with the user annotations?
Good point. We can make it IDENTIFIER:WIDTH:TAG, if that makes more sense to you. TAG could be a number or colour.
Download a gene and it normally gets assigned to tag group one, black. The default would be a whisker length of 100. If you want to add genes, or re-assign tags of loaded genes, or change the whisker length, you can use the annotation box.
Would you like to Skype tomorrow?
On 2013-10-10, at 9:40 PM, Hans Yu notifications@github.com wrote:
I've already wrote a class for user annotations. Now I just need to confirm whether the annotations would be created using the IDENTIFIER:WIDTH:COLOR notation or something else (like the color choices on annotation popup). I plan to combine the annotation system with marking genes by the Get Data button, i.e. genes annotated will be loaded just like hitting the Get Data button for that gene. I really like the idea of having the color choices on the annotation popup, but how can we make it consistent with the user annotations?
— Reply to this email directly or view it on GitHub.
OK that would be a good way to go. I'm working the grocery job tomorrow, unfortunately. By the time I come back it'll be 8pm in Toronto. Can you Skype during the long weekend?
I'm going to a data vis conference in Atlanta on Saturday evening through Thursday. I'm meeting Nick there. Maybe we can all talk on Sunday or Monday?
On Oct 10, 2013, at 11:47 PM, Hans Yu notifications@github.com wrote:
OK that would be a good way to go. I'm working the grocery job tomorrow, unfortunately. By the time I come back it'll be 8pm in Toronto. Can you Skype during the long weekend?
— Reply to this email directly or view it on GitHub.
Either day is fine with me. Just let me know the time! On Oct 11, 2013 7:03 AM, "Jamie Waese" notifications@github.com wrote:
I'm going to a data vis conference in Atlanta on Saturday evening through Thursday. I'm meeting Nick there. Maybe we can all talk on Sunday or Monday?
On Oct 10, 2013, at 11:47 PM, Hans Yu notifications@github.com wrote:
OK that would be a good way to go. I'm working the grocery job tomorrow, unfortunately. By the time I come back it'll be 8pm in Toronto. Can you Skype during the long weekend?
— Reply to this email directly or view it on GitHub.
— Reply to this email directly or view it on GitHub.
I've added a simplified version for user annotation. Right now it only supports the notation AGI:WIDTH:COLOR. A valid input is an existing agi for AGI, a number for WIDTH, and any string for COLOR. I still have to implement COLOR/TAG so right now I am not checking whether COLOR is a valid hexadecimal.
Tomorrow, I'll try to implement annotation popup when mouse is over the marked gene's label. It is going to be tricky because the labels are drawn as Processing texts (so they are Canvas texts). As you know, there is no automatic tracking for them so I'll have to write a wrapper class for texts which can track them.
Implemented in the uncleaned code, so I'm closing this. The cleaned up version will have this feature working once I can retrieve genes dynamically via web service, instead of loading all genes at initialization.
One feature Nick mentioned to me last week was that the whisker shouldn't be centered over the chromosome. Rather, it should stick out to the left if the gene direction is positive (a + in the GFF file), and to the right if the gene direction is negative (a - in the GFF file).
We should also put a maximum size in effect. No need for us to allow a line halfway across the page.
On Oct 20, 2013, at 8:35 PM, Hans Yu notifications@github.com wrote:
Implemented in the uncleaned code, so I'm closing this. The cleaned up version will have this feature working once I can retrieve genes dynamically via web service, instead of loading all genes at initialization.
— Reply to this email directly or view it on GitHub.
As per Nick's request.
Entry might be: AT1G19205:25:#FF0000, HYDRA1:50:#99CC00, etc. (gene name):(width of whisker in number of pixels):(color). Comma separated entries, unlimited size.
One issue with using pixels for width of the whiskers is that they won't resize when we zoom in. Perhaps the width should actually be a percentage from 1 to 100, and we can resize according to the min and max of the current zoom level.