zjuvai / drgraph Goto Github PK

Hi,

first of all, congratulations for this library and for your manuscript. It is very interesting and I'm eager to test it into some applications.

Yet, it is unclear to me how exactly to use the library. Is there any further documentation? Furthermore, is it possible to use the library entirely from python, not necessarily using in-disk data?

Thanks

run the code:
./Vis -input ./data/block_2000.txt -output block_2000.txt -neg 10 -samples 400 -gamma 0.1 -mode 1 -A 2 -B 1
but get:

[INPUT FILE] ./data/block_2000.txt    
[LABEL FILE] 
[OUTPUT FILE] block_2000.txt                                                                                            
[KNN FILE]                                                                                                              
disMax: 1 n_samples: 400 n_gamma: 0.100 n_negative: 10                                                                  
Reading graph edges from ./data/block_2000.txt                                                                          
Reading 15K edges                                                                                                       
Segmentation fault

Has the same environment as described :
Gsl(2.4) Boost(1.58) Cmake(3.11)

Backend installation uses a requirements.txt file, which is not present.

When I ran build.sh, I initially ran into a problem saying "reference to 'real' is ambiguous.I encountered another problem after changing 'real' to 'float' in the file. The problem is:
./src/algorithm/kdtreeub_index.hpp:1192:7: require from here
I don't understand what this means. I hope I can get your help.