From Protein Prediction 2 Winter Semester 2014
Revision as of 10:48, 7 January 2015 by Ppwikiuser (talk | contribs) (→Week 5(19.12.2014 - 25.12.2014))
- 1 Git Repository
- 1.1 1. GUI mockups
- 1.2 How we are going to improve it ?
- 1.3 2. Application design
- 1.4 3. Data Format
- 1.5 4. Roadmap
- 1.6 Week 1(21.11.2014 - 27.11.2014)
- 1.7 Week 2(28.11.2014 - 4.12.2014)
- 1.8 Week 3(5.12.2014 - 11.12.2014)
- 1.9 Week 4(12.12.2014 - 18.12.2014)
- 1.10 Week 5(19.12.2014 - 25.12.2014)
- 1.11 Week 6(26.12.2014 - 1.1.2015)
- 1.12 Week 7(2.1.2015 - 8.1.2015)
|Week number||Completed tasks||Status|
1. GUI mockups
- Currently the user interface looks like this
How we are going to improve it ?
By incorporating the following visualizations
- 3D Globe for Maps
- Word Cloud for filtering.
- Bilevel/Zoomable Partition Layout for Events.
3D Globe for maps
- Can able to Rotate the map.
- Able to view information of each events, here we plan to implement this module using the technique Orthographic to Equirectangular.
- Along with the event information corresponding url will also be embedded.
- User can filter the events based on country and date.
- Interconnected lines between the countries which are likely to have events happening on same date.
BiLevel/Zoomable Partition Layout for Events
- Navigation through events, with an emphasis on usability.
|Zoomable Partition||BiLevel Partition|
- Each inner circle is representing root category, i.e. country, which expands on click and displays details, i.e. events in a selected country.
Word Cloud for Keywords
- Visualization of the most used keywords in iAnn search.
- Based on popularity, the size of each word varies.
2. Application design
Expected technical difficulties
- Implementing Orthographic to Equirectangular on each marker points(countries).
- Interconnecting lines between the countries when the events from different countries likely to happen on same date, because we need latitude and longitude points to map.
Fancy libraries you plan to use
3. Data Format
- Input is likely to be in json format with events,location and date, etc.
- With the above information it’s difficult to map the exact location in map, we need the latitude and longitude points to overcome it.
- Finalize our proposed ideas.
- Developing all the required and accepted visualization components.
- Finally Integrating into widget and upload to BioJs
Week 1(21.11.2014 - 27.11.2014)
"Mentors are happy with our proposed ideas"
After the discussion with our mentors we were able to finalize our task.
Below you can find the list of tasks based on the priority
- Word Cloud
- 3D Globe
- Bilevel Partition (if time permits)
- Configurating Tomcat server instead of Jetty, due to security concerns (if time permits)
- Preparing endpoints for our visualisation widgets (if time permits)
- Understanding the working of Ajax Solr.
- Integrating the new components with the existing widget.
- Setup of repositories, working directories and shared documents
- Go through some basic tutorials of Ajax Solr.
- We were successfully able to design a word cloud using D3 for one of the filter options "Host", but need to work on size and onclick events.
Plans for Week 2
- Implement the word cloud for rest of the filter options.
- And also to include the onclick events.
- Initial setup for 3D Globe
- Deliver proposal (document) for changes in a Solr schema and discuss them with project sponsors.
Week 2(28.11.2014 - 4.12.2014)
- Implementation of Word Cloud for advanced filter options is done, waiting for the feedback.
- Greater the size of word, the more popular it is.
GitHub : iANN_WORD_CLOUD
- Initial setup for 3D Globe is done.
Note : Currently there is a problem in running with chrome due to queue library for asynchronous request, it is advised to view either in Mozilla firefox or safari
- In addition filter based on country is also included.
- List of all countries are available in dropdown for user selection.
- Rotate function is also incorporated.
- Based on the country selection, the globe rotates and points to a specific location.
- Transformation of 3D to 2D(Orthographic to Equirectangular) is also completed.
- By clicking on specific country on the globe, the 3D to 2D transformation happens and the selection area remains highlighted.
- The reverse process i.e 2D to 3D can be achieved by clicking again on the layout.
GitHub : iANN_3D_GLOBE
Week 3(5.12.2014 - 11.12.2014)
- We have improved our word cloud look compared to our previous week.
- Also we had an issue in bringing all the words in to cloud and now this is also solved.
- We tried with adding markups in the Globe, but we had an issue while converting to 2D map, as you can imagine the markup positions varies for each country with respect to 3D and 2D view, that transition we can't able to achieve with the markup.
- We even tried with adding colors for each country, but it doesn't looks good.
- We have developed the Tree map with respect to keywords.
- Currently it's not interactive, we are working on it.
- Once we are done with Tree map, we will also do the BiLevel Partition which is one of the visualization our mentors are very much interested.
Week 4(12.12.2014 - 18.12.2014)
- With the addition of links and hover functionality word cloud comes to an end.
- When you move the cursor over the word cloud the words gets zoom out a bit to show the user that it is clickable.
- Now in the word cloud, each word is incorporated with search functionality. For example : if United Kingdom is clicked the remaining three fields will display the words related to it.
- Finally we were able to solve the challenge of adding markers to the 3D rotating Globe.
- Markers are also incorporated for 2D view.
- Tooltip is included as a mouse over event, so that it displays the country name according the movement of cursor.
- Finalizing the hierarchy structure to be used in the partition layout.
- Working on the input data inorder to convert it into JSON format.
Week 5(19.12.2014 - 25.12.2014)
- Integrated with iAnn viewer.
- Initially the transition happens when we double click on all the country, now we have made sure that the transitions happens only for the country which has the events.
- Automatic zoom in functionality is added in the 2D view. i.e Once when the user clicks the interested country on globe the transition happens followed by that selected country alone gets zoomed in automatically.
- When the selected country gets zoomed in, only the markers pertaining to that country will be displayed.
- Following that List of all events w.r.t selected country will get displayed at the bottom.
- In order to make it flexible for non-savvy users the option of selecting 3D or 2D view is provided.
- Stabilizing the transition between 3D and 2D view.
Week 6(26.12.2014 - 1.1.2015)
- The Hierarchy for partition was finalized such that : Country->City->Events.
- Both Bilevel and Treemap depicts the same concept, but anyways we have implemented both.
- In Treemap, all hierarchy will be visible in the beginning, but on click of each cell, the selected area will get zoomed in and displays only the cells related to it.
- In Bilevel also same concept is applied, on click of each area detailed view can be obtained.
Week 7(2.1.2015 - 8.1.2015)
Enhancements and Fixes in 3D Globe
- The Rotation speed of the globe is increased.
- In 3D view, when the country gets selected from dropdown, the corresponding place gets highlighted and zoomed in automatically, previously it was happening only in 2D.
- Some features were missing in 2D, but now we have make sure that both 2D and 3D view has same features in a way which is convenient for all users.
- Work in progress for making it as individual and reusable compenet for BioJs.