Thanks for your quick response.
Before starting to code, you must read "Gutwenger, Carsten; Mutzel, Petra > (2001), "A linear time implementation of SPQR-trees", Proc. 8th > International Symposium on Graph Drawing (GD 2000), Lecture Notes in > Computer Science, 1984, Springer-Verlag, pp. 77–90, > doi:10.1007/3-540-44541-2_8" > It corrects the Hopcroft-Tarjan algorithm. I have completed reading the research paper[2] "Gutwenger, Carsten; Mutzel, Petra (2001), A linear-time implementation of SPQR-trees", which I have mentioned in my first mail. All the corrections suggested by paper[2] will be incorporated in the code. Why using an extra class array_info instead of including all arrays in the > class Search ? This way you can initialize the arrays only if you need them. Thanks for the suggestion yes, I will merge array_info to Search class. In your code, please add sufficient amount of comments. This is very > important. You can look at the code of the graph module of sagemath. You > will see that the amount of comments is huge. I have gone through most of the graph module of Sage(modular decomposition, graph etc), yes there have a good amount of comments for understanding, I will try to add all the sufficient comments which require understanding the module. Shall I go ahead and complete the implementation of Triconnected components? Thanks, Harsh On Saturday, March 3, 2018 at 3:19:24 PM UTC+5:30, david....@gmail.com wrote: > > > > Le vendredi 2 mars 2018 22:55:27 UTC+1, Sai Harsh Tondomker a écrit : >> >> Thanks for your response. I would like to go with recoding the method >> with documentation. >> >> I have installed OGDF and sage on my system, using OGDF implementation as >> a reference, I am implementing Find Triconnected Components >> <http://www.dtic.mil/dtic/tr/fulltext/u2/746856.pdf> in a graph for >> sage. Code link is given below. >> > > Before starting to code, you must read "Gutwenger, Carsten; Mutzel, Petra > (2001), "A linear time implementation of SPQR-trees", Proc. 8th > International Symposium on Graph Drawing (GD 2000), Lecture Notes in > Computer Science, 1984, Springer-Verlag, pp. 77–90, > doi:10.1007/3-540-44541-2_8" > > It corrects the Hopcroft-Tarjan algorithm. > > > >> >> I have been working on an implementation of the Triconnected components >> algorithm in the research paper in python. The algorithm requires a lot of >> DFS function calls, corresponding to it. I made some design decisions which >> I needed to get reviewed. >> >> 1. I have stored all the arrays in a class (array_info) such that >> I can access them in any other classes/functions. >> >> a. The advantage of this approach is that there is no >> need to send all the parameters(arrays) to the functions again and again. >> >> b. The demerit is that if a function needs only a few >> lists from array_info, all the lists in the array_info will be initialized. >> > > Why using an extra class array_info instead of including all arrays in the > class Search ? This way you can initialize the arrays only if you need them. > > >> >> 2. For DFS to avoid passing graph at each function call, I stored >> the graph in a parameter of a class (search). >> > > OK > > >> >> I am currently implementing the remaining code required for finding >> triconnected components. I needed some review of these design decisions and >> suggestions if any changes I needed to make for better performance. >> > > In your code, please add sufficient amount of comments. This is very > important. You can look at the code of the graph module of sagemath. You > will see that the amount of comments is huge. > > >> >> Code: link <https://github.com/SaiHarsh/Triconnected-Component> >> >> Thanks, >> Harsh >> >> >> -- You received this message because you are subscribed to the Google Groups "sage-gsoc" group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-gsoc+unsubscr...@googlegroups.com. To post to this group, send email to sage-gsoc@googlegroups.com. Visit this group at https://groups.google.com/group/sage-gsoc. For more options, visit https://groups.google.com/d/optout.
