cmake error in ply_c #138

shobhitjdpsr · 2019-07-24T23:52:18Z

While running cmake . -DPYTHON_LIBRARY=$CONDAENV/lib/libpython3.6m.so -DPYTHON_INCLUDE_DIR=$CONDAENV/include/python3.6m -DBOOST_INCLUDEDIR=$CONDAENV/include -DEIGEN3_INCLUDE_DIR=$CONDAENV/include/eigen3

I am getting an error that boost is too old, despite of installing boost 1.66.

What all can I do to make this work?

shobhitjdpsr · 2019-07-25T04:20:02Z

@loicland

loicland · 2019-07-25T10:18:49Z

Hi,

what did your set CONDAENV to?

and what is the results of

conda list

in a terminal?

shobhitjdpsr · 2019-07-25T15:27:39Z

@loicland Please have a llook

shobhitjdpsr · 2019-07-25T16:08:20Z

@Pandinosaurus @loicland

loicland · 2019-07-25T18:34:22Z

CONDAENV should be more like something like
/home/your_session/anaconda3

Find it by typing
locate anaconda

shobhitjdpsr · 2019-07-25T19:03:22Z

@loicland
Never mind I got it working., by removing cmakecache.txt .

But next error is in make

shobhitjdpsr · 2019-07-25T19:04:43Z

In ply_c while doing MAKE, I am getting following error:

[ 50%] Building CXX object CMakeFiles/ply_c.dir/ply_c.cpp.o
/home/spartan/Desktop/superpoint_graph/partition/ply_c/ply_c.cpp:4:23: fatal error: Eigen/Dense: No such file or directory
#include <Eigen/Dense>
^
compilation terminated.
CMakeFiles/ply_c.dir/build.make:62: recipe for target 'CMakeFiles/ply_c.dir/ply_c.cpp.o' failed
make[2]: *** [CMakeFiles/ply_c.dir/ply_c.cpp.o] Error 1
CMakeFiles/Makefile2:67: recipe for target 'CMakeFiles/ply_c.dir/all' failed
make[1]: *** [CMakeFiles/ply_c.dir/all] Error 2
Makefile:83: recipe for target 'all' failed
make: *** [all] Error 2
(my_env) spartan@spartan-

loicland · 2019-07-25T19:48:00Z

Have you tried the latest commit? Otherwise reinstall Eigen.

shobhitjdpsr · 2019-07-25T19:49:07Z

I tried the commit which was there 24 hours ago approx

@loicland

shobhitjdpsr · 2019-07-25T20:22:19Z

@loicland
Nevermind, This was fixed by editing partition/ply_c/ply_c.cpp and changing TO:-
#include <eigen3/Eigen/Dense>
#include <eigen3/Eigen/Eigenvalues

**BUT, NOW I AM GETTING FOLLOWING ISSUE:-

fatal error: pyconfig.h: No such file or directory**

loicland · 2019-07-25T21:11:36Z

Check whether your anaconda is using python3.6m, and if not adapt the cmake options (and dont forget to destroy all the cmakefiles).

shobhitjdpsr · 2019-07-25T21:20:20Z

Thanks, it worked, I changed to 3.7m in cmake, because thats what I had in my cmake

Now @loicland

cmake .. -DPYTHON_LIBRARY=$CONDAENV/lib/libpython3.7m.so -DPYTHON_INCLUDE_DIR=$CONDAENV/include/python3.7m -DBOOST_INCLUDEDIR=$CONDAENV/include -DEIGEN3_INCLUDE_DIR=$CONDAENV/include/eigen3

I am getting error

shobhitjdpsr · 2019-07-25T23:16:41Z

@loicland

Now I will summarize the whole problem that I am facing currently:-

While compiling libcp, and running cmake (cmake .. -DPYTHON_LIBRARY=$/home/spartan/anaconda3/lib/libpython3.7m.so -DPYTHON_INCLUDE_DIR=$/home/spartan/anaconda3/include/python3.7m -DBOOST_INCLUDEDIR=$/usr/local/lib -DEIGEN3_INCLUDE_DIR=$/home/spartan/anaconda3/include/eigen3
)
Note:- I have python3.7m in my anaconda3/lib and it worked for ply_c

Now , I am getting the following errors:-

ALso while running locate libboost python, I am getting following errors:-

loicland · 2019-07-26T08:28:16Z

you should remove the $ before your anaconda path.

shobhitjdpsr · 2019-07-26T15:41:35Z

@loicland
Even after doing the same, I am still getting errors

shobhitjdpsr · 2019-07-26T16:09:47Z

loicland · 2019-07-26T16:51:15Z

Have you tried the symbolic link fix I added today to the readme?

Otherwise replace in cut pursuit /sec/cmakeist
numpy${PYTHONVERSION}
By
numpy

shobhitjdpsr · 2019-07-26T16:57:03Z

Do you mean just the line highlighted to be changed?

shobhitjdpsr · 2019-07-26T16:59:40Z

@loicland

shobhitjdpsr · 2019-07-26T17:03:45Z

@loicland
I did try the new command in readme for symbolic link, but still I am getting and error. Should I change the cmakeLists also?

loicland · 2019-07-26T17:11:40Z

Anywhere bumpy appears. But try the symbolic link first.

And dont forget to remove makefiles etc... after the changes.

shobhitjdpsr · 2019-07-26T17:21:42Z

@loicland
Finally, cmake runs with a warning.
[
CMake Warning:
Manually-specified variables were not used by the project:

EIGEN3_INCLUDE_DIR

-- Build files have been written to: /home/spartan/Desktop/superpoint_graph/partition/cut-pursuit/build
]

But now make has some issues:-

include <pyconfig.h>

                   ^compilation terminated.

src/CMakeFiles/cp.dir/build.make:62: recipe for target 'src/CMakeFiles/cp.dir/cutpursuit.cpp.o' failed
make[2]: *** [src/CMakeFiles/cp.dir/cutpursuit.cpp.o] Error 1
CMakeFiles/Makefile2:85: recipe for target 'src/CMakeFiles/cp.dir/all' failed
make[1]: *** [src/CMakeFiles/cp.dir/all] Error 2
Makefile:83: recipe for target 'all' failed
make: *** [all] Error 2

loicland · 2019-07-26T17:36:59Z

This thread adapted to your python version.

shobhitjdpsr · 2019-07-26T18:49:19Z

@loicland

No again a warning has popped up.
Please have a look

shobhitjdpsr · 2019-07-26T19:05:52Z

@loicland

I tried various other things to export path, but issue persists....

Please have a look

shobhitjdpsr · 2019-07-26T19:17:38Z

@loicland : Have a look at this too

shobhitjdpsr · 2019-07-26T20:13:21Z

@loicland

While doing cmake, I was getting following cmake-warning.

Submitting for your reference:-

loicland · 2019-07-26T21:05:53Z

Isn't your python version 3.7m? Try the export with this, and dont forget to erase the cmakefiles before.

shobhitjdpsr · 2019-07-26T21:13:38Z

@loicland : What do you think might be causing the same and what can I do to rectify it

For your reference:-
ALso I already have libbpython3.7m.so by default in this location:-

shobhitjdpsr · 2019-07-26T21:25:40Z

shobhitjdpsr · 2019-07-26T21:26:19Z

@loicland
I am using 3.7m in export, but the error persists

loicland · 2019-07-26T21:28:30Z

Your path has an upper case to the first letter of Anaconda which is not the case in your drive.

loicland · 2019-07-26T21:41:40Z

Yep, it's done. Choose a dataset, open its .md file and follow the instructions.

Or run the test.py function of cut pursuit.

shobhitjdpsr · 2019-07-29T22:20:17Z

@loicland :
In order to run the highlighted pretrained model, do we necessarily need CUDA?

I tried putting pretrained models in results/s3dis/pretrained/best after extracting like this:_

Then, I put test.sh in Desktop/superpoint_graph like this:-

Then after running this on terminal I get this:-

@loicland
Am I doing it the right way? What to do with the errors?
Also do we necessarily need CUDA installed? aSking so as its not mentioned in requirements

shobhitjdpsr · 2019-07-29T22:35:37Z

@loicland

On running ./test.sh ,

I get something like this:-

original error: libcublas.so.9.0: cannot open shared object file: No such file or directory

Some more screenshot for your reference:-

loicland · 2019-07-30T08:17:34Z

you can use the option --cuda 0

Also it seems that cupy is not found. Try to install it through conda:

 conda install -c anaconda cupy

nicolas-chaulet · 2019-07-30T11:26:22Z

You won't be able to install cupy if you don't have the nvidia drivers or no GPU, which is fine. With the flag above it should all work.

shobhitjdpsr · 2019-07-30T16:49:51Z

@loicland @nicolas-chaulet
I tried installing using
conda install -c anaconda cupy

Then I updated test.sh, located here:-

I tried testing with test.sh containing all this:-

Bit I get same error:-

On doing pip freeze I get this:-

I tried removing the one I did try running through through conda's cupy but I get error when running test.sh:-

Also, when I tried running through only cupy-cuda90==6.2.0 installed, I was still getting errors.(same has been attached in my previous yesterday's message)

@loicland @nicolas-chaulet
What would you recommend now?

loicland · 2019-07-30T17:00:18Z

what are the output of:
nvcc --version
and
conda list
in a terminal (in text form please, more readable than screenshot of terminals)

shobhitjdpsr · 2019-07-30T17:02:58Z

@loicland

On running nvcc --version, I get:-
The program 'nvcc' is currently not installed. You can install it by typing:
sudo apt install nvidia-cuda-toolkit

Also, conda list gives the following:-

WARNING: The conda.compat module is deprecated and will be removed in a future release.

packages in environment at /home/spartan/anaconda3/envs/my_env:

loicland · 2019-07-31T08:38:42Z

Hi,

pull a fresh commit, it seems like your cuda_kernels.py is outdated.

@nicolas-chaulet I see you added a pass when cupy fail to import, but doesn't it cause an error line 42 of cuda_kernels.py when the import fails?

I don't see pytroch in either your pip or conda install. In a terminal, run:

conda install -c pytorch pytorch

shobhitjdpsr · 2019-07-31T23:03:19Z

@loicland @nicolas-chaulet

When trying to run test.sh containing:-

I get the following error:-
Traceback (most recent call last):
File "learning/main.py", line 444, in
main()
File "learning/main.py", line 159, in main
train_dataset, test_dataset, valid_dataset, scaler = create_dataset(args)
File "/home/spartan/Desktop/SJ/superpoint_graph/learning/s3dis_dataset.py", line 55, in get_datasets
trainlist, testlist, validlist, scaler = spg.scaler01(trainlist, testlist, validlist=validlist)
File "/home/spartan/Desktop/SJ/superpoint_graph/learning/spg.py", line 53, in scaler01
edge_feats = np.concatenate([ trainlist[i][3] for i in range(len(trainlist)) ], 0)
ValueError: need at least one array to concatenate

Also, I tried searching in s3dis.py, it is searching .h5 files, but there are no .h5 files in stanford dataset, see:-

loicland · 2019-08-01T08:19:47Z

Your S3DIS_DIR folder should have a /data folder, a /featuresfolder, a superpoint_graphs folder and a /parsed folder. All of these contains folder by area, and inside a .h5 file per room. Follow the instructions (partition and parsing) and you should get there.

whuhxb · 2023-02-13T13:56:17Z

In ply_c while doing MAKE, I am getting following error:

[ 50%] Building CXX object CMakeFiles/ply_c.dir/ply_c.cpp.o /home/spartan/Desktop/superpoint_graph/partition/ply_c/ply_c.cpp:4:23: fatal error: Eigen/Dense: No such file or directory #include <Eigen/Dense> ^ compilation terminated. CMakeFiles/ply_c.dir/build.make:62: recipe for target 'CMakeFiles/ply_c.dir/ply_c.cpp.o' failed make[2]: *** [CMakeFiles/ply_c.dir/ply_c.cpp.o] Error 1 CMakeFiles/Makefile2:67: recipe for target 'CMakeFiles/ply_c.dir/all' failed make[1]: *** [CMakeFiles/ply_c.dir/all] Error 2 Makefile:83: recipe for target 'all' failed make: *** [all] Error 2 (my_env) spartan@spartan-

How to solve this compilation error? Thanks.

shobhitjdpsr closed this as completed Jul 29, 2019

shobhitjdpsr reopened this Jul 29, 2019

cmake error in ply_c #138

cmake error in ply_c #138

Comments

shobhitjdpsr commented Jul 24, 2019

shobhitjdpsr commented Jul 25, 2019

loicland commented Jul 25, 2019 • edited Loading

shobhitjdpsr commented Jul 25, 2019

shobhitjdpsr commented Jul 25, 2019

loicland commented Jul 25, 2019 • edited Loading

shobhitjdpsr commented Jul 25, 2019

shobhitjdpsr commented Jul 25, 2019

loicland commented Jul 25, 2019

shobhitjdpsr commented Jul 25, 2019

shobhitjdpsr commented Jul 25, 2019 • edited Loading

loicland commented Jul 25, 2019

shobhitjdpsr commented Jul 25, 2019 • edited Loading

shobhitjdpsr commented Jul 25, 2019

loicland commented Jul 26, 2019

shobhitjdpsr commented Jul 26, 2019

shobhitjdpsr commented Jul 26, 2019

loicland commented Jul 26, 2019 • edited Loading

shobhitjdpsr commented Jul 26, 2019

shobhitjdpsr commented Jul 26, 2019

shobhitjdpsr commented Jul 26, 2019

loicland commented Jul 26, 2019 • edited Loading

shobhitjdpsr commented Jul 26, 2019 • edited Loading

include <pyconfig.h>

loicland commented Jul 26, 2019

shobhitjdpsr commented Jul 26, 2019

shobhitjdpsr commented Jul 26, 2019

shobhitjdpsr commented Jul 26, 2019

shobhitjdpsr commented Jul 26, 2019

loicland commented Jul 26, 2019

shobhitjdpsr commented Jul 26, 2019 • edited Loading

shobhitjdpsr commented Jul 26, 2019

shobhitjdpsr commented Jul 26, 2019

loicland commented Jul 26, 2019

loicland commented Jul 26, 2019 • edited Loading

shobhitjdpsr commented Jul 29, 2019

shobhitjdpsr commented Jul 29, 2019

loicland commented Jul 30, 2019

nicolas-chaulet commented Jul 30, 2019

shobhitjdpsr commented Jul 30, 2019

loicland commented Jul 30, 2019

shobhitjdpsr commented Jul 30, 2019 • edited Loading

packages in environment at /home/spartan/anaconda3/envs/my_env:

Name Version Build Channel

loicland commented Jul 31, 2019 • edited Loading

shobhitjdpsr commented Jul 31, 2019 • edited Loading

loicland commented Aug 1, 2019

whuhxb commented Feb 13, 2023

loicland commented Jul 25, 2019 •

edited

Loading

loicland commented Jul 25, 2019 •

edited

Loading

shobhitjdpsr commented Jul 25, 2019 •

edited

Loading

shobhitjdpsr commented Jul 25, 2019 •

edited

Loading

loicland commented Jul 26, 2019 •

edited

Loading

loicland commented Jul 26, 2019 •

edited

Loading

shobhitjdpsr commented Jul 26, 2019 •

edited

Loading

shobhitjdpsr commented Jul 26, 2019 •

edited

Loading

loicland commented Jul 26, 2019 •

edited

Loading

shobhitjdpsr commented Jul 30, 2019 •

edited

Loading

loicland commented Jul 31, 2019 •

edited

Loading

shobhitjdpsr commented Jul 31, 2019 •

edited

Loading