As discussed with @drphilmarshall during the HackWeek at CMU, we plan to connect Hoopla to some Professional lens modeling codes too. The goal is to improve the accuracy of the lens model in Hoopla using the codes like: Lensed, Glafic, Gravlens, Lenstool...etc. However, calling an executable in javascript is not straightforward because of the security policy in JavaScript. Therefore, I tried to implement this using PyQT. Some progress is shown below.
Created a basic UI with sliders to guess original model (the Red Rectangle) and a button to call optimization function (the Purple Rectangle).
Embedded an optimization function (Scipy.optimize for the first test) in the code to optimize the starting model. The optimized model is visualized in the below picture. Compared to the image above, the model is clearly more accurate.
I will create another repo for this issue and construct this interface as a universal graphical user interface for the lensing codes mentioned above so that lens modelers can model simple lensing systems interactively adopting different pipelines.
linan7788626
commented
7 months ago
As discussed with @drphilmarshall during the HackWeek at CMU, we plan to connect Hoopla to some Professional lens modeling codes too. The goal is to improve the accuracy of the lens model in Hoopla using the codes like: Lensed, Glafic, Gravlens, Lenstool...etc. However, calling an executable in javascript is not straightforward because of the security policy in JavaScript. Therefore, I tried to implement this using PyQT. Some progress is shown below.