Fn/eventual fixing lo sand sub #88
Conversation
…ventualFixingLoSandSub
… the moment implemented only for substructure. It does not work so well and it needs to be developed further.
…ventualFixingLoSandSub
…ventualFixingLoSandSub
…ualFixingLoSandSub
…ventualFixingLoSandSub
… into fn/EventualFixingLoSandSub
…ventualFixingLoSandSub
…nformations on the substructure to test it.
…field planes than the substructure plane.
…er field planes than the substructure plane.
…s some of the properties of the substructure. Also several changes were made in Lens::insertSubstructures, Lens::resetSubstructure, and Lens::deleteSubstructures in order to avoid some errors happening when there is no substructure halos.
…hat I can use them in my driver program. Also noticed that before rho of substructure was not actually saved.
…ly when there is no parameter file.
…sHalo to get out of sync with each other.
… the substructure halos in insertSubstructures and resetSubstructure. Corrected a typo with the lengths[] test in TreeQuad::BuildQTreeNB.
…the parameter file halos and not getting the error from assert(Dist == -1).
…d Lens::replaceMain(). Also imposed some more discipline on the LenHalo hierarchy by making zlens, mass and Rsize private variables of LensHalo and do the assignment of these variables from the parameter file through the LensHalo constructor only.
…ventualFixingLoSandSub
|
I think we should do the merge pull request now as everything seems to work fine on SLsimLib side. |
|
The problem is that when I diff the head of your branch with master there are a I think the best way to proceed is for you to find the commits on your branch and |
|
Okay. So if I understand well I put myself on the master branch, then I create a new branch and cherry pick the commits one by one according to their "meaning", but not particularly according to their date, right ? What should I do with the commits which consisted in merging the master branch into my own branch (and eventually fixing conflicts) ? |
|
I am also wondering: which is the first commit I should do the cherry-pick on ? Dec 1, 2015, right ? The commits before where already merged I think. |
|
Hi Fabien, You should do the cherry picking in chronological order. Otherwise you might The priority is on getting the paper done though. |
…ventualFixingLoSandSub
|
Hi Ben, |
|
I did not start the cherry pick from the moment the branch (fn/EventualFixingLoSandSub) was created though, but instead from the last time it was merged into the master branch. I guess this is not a problem. |
|
Lets concentrate on the research and worry about this later. |
|
I solved the problems I think and it seems to work with my project now. I am running it over the different lenses in order to check with previous runs I had done and see if anything is affected. Maybe we can merge the cherry-picked branch right after I have proceeded with a visual check of my results (compared with previous runs) ? Hence this problem will be solved and we can focus on the physics indeed :). |
|
Okay I checked that the results are exactly the same between now (on my cherry pick branch) and previously on my local branch. I had merged the master branch into my cherry pick branch this morning and there was no conflict. Should I do a pull request of my cherry-pick branch ? |
|
Yes, that would be more easy to review. |
The main contribution in this branch concerns the substructure. I have rewritten the organization of the functions insert/reset/deleteSubstructure(s) () and tested them in many different configurations (including also low numberdensity that generates no halos). I merged also with the master branch and solved conflicts. I should review all the changes but I think it will not create problems.