xbldev / datavisualization

Data visualization exercise, for now with D3.js and Python(matplotlib, seaborn etc)

DataVisualization

Data visualization exercise, for now with D3.js and Python(matplotlib, seaborn etc)

Comment 30/11/2017, 3:42 pm:

Added one responsive page which uses Bootstrap styling, with nav-bar on top which contains menu items and will shrink into a button on the right when the screen is too small, and 12 col grid in the middle with 2 col wide links on the left, 8 col wide center area, and 2 col wide well class menu list on the right.

Comment 21/11/2017, 8:10 pm:

Added more Mocha tests based on the original tests in thte tutorial: https://www.meteor.com/tutorials/angular/testing

For now it seems that it won't be easy, or even possible, to write tests that can test front + back end, just like when testing ReactJS/Redux + NodeJS with Jest + Enzyme, for now the input to backend can only be simulated/hardcoded, and for the front end in the todo-list the tasks contained in the controller also have to be hard-coded.

The aspects that can test with backend: mostly just function calls, but because some meteor built-in functions such as meteor.userId() cannot be called when doing the test, mock function for each real function is needed to run the test, the error thrown for bad result can be tested too.

The asepcts that can test with frontend: certain UI element values, such as the value for a checkbox(on/off), value for an input field, or the number of li elements contained in the ul element; controller and its function calls, mostly just checking if a meteor method is called and if it's called with the appropriate parameters provided by the frontend.

Comment 21/11/2017, 2:58 pm:

Added D3/Angular_Meteor_D3, basically a combination of: 1. the beginner angular2 + meteor tutorial from: https://www.meteor.com/tutorials/angular/creating-an-app, 2. AngularJS routing with angular-route, 3. Some example D3 data visulisation from: https://bl.ocks.org/.

The idea is that for each different page there's a different routing, and with each routing there's a different template, which is a usually just a html file with svg or div, and a controller, which is the D3 JS script that controls what the the template renders.

For now problems include: some function calls are only available in 1 version of D3 but not in others; layout of the rendering often goes out of the container div; some examples from the website contain JS files that generate errors.