Weekly Report 22Oct2010
From RHESSI Wiki
Documentation for Users wishing to add content to the new RHESSI Website, Migration of Content to the new site
Created a document explaining how to edit content in an simple manner on the new RHESSI Website. The document explains how to create a new page, add rich text content, including links, and add images to pages. It also links to external documents which have a more in-depth explanation of editing content using the typo3 Content Management System.
Content has been migrated by users to the draft workspace and is being published to the live site upon review. To view the live site check here. Though it is not currently ready for public use, we are making progress. I think the end of November is a reasonable time line to have the site ready for public use.
Annapolis Image Test of algorithms
I've modified the scripts written by Richard Schwartz to create images using multiple simulated data files, returning the images, along with statistics comparing the images created to the original data maps. The modified script can be found here. Using this script I've created images using the Clean algorithm and the forward fit algorithm (not using visibilities). These images have have been saved in a .sav file, after I show them to Brian they will be uploaded to the RHESSI Imaging test site in Genoa.
As an example of the images created the following was made with the Clean algorithm. The original map is shown in white contours, the reconstructed Clean image is the background. All detectors were used, and the clean_regress_combine factor was set.
The C statistics for this image are shown below. The plot was made using hsi_image_profile_plot.pro.
Significant time was spent tracking down why there was a difference the C statistic, calculated by calling c_statistic.pro directly from the script and calculating it using hsi_image_profile_plot.pro. I got around this by calling the c_statistic routine for each detector individually using a loop and it solved the issue.
Goals for next week
- Create clean and forward fit images for the image test using best possible parameters for each algorithm. This should show how close the algorithms can get to reproducing a know map.
- Add extensions to the typo3 backend of the website to allow for filelist additions and allow the site to update information such as the RHESSI Science nugget automatically. Check the first draft of the website documentation, and add parts on the extensions.
- Arrange a meeting with Brian, Eduard, and myself to discuss the Pixon albedo imaging studies, and decide how to progress with the study.
- Continue migrating content to the new website
