Learning Feasibility

This repo has code for discovering object-level latent spaces based on interactions with that object. This is an extension of the stacking repository and Tmuch of the robot functionality comes from the rachelholladay/pb_robot repository.

Installation

Dependencies: python3, git

1. Install latent_feasibility
   1. Clone this repository.
   2. Create a virual environment for this project.
      1. cd latent_feasibility
      2. virtualenv .venv --python=/usr/bin/python3
      3. source .venv/bin/activate
   3. Install dependencies.
      1. xargs -n1 pip3 install < requirements.txt
      2. We are currently using PyTorch 1.9.1: pip install torch==1.9.1+cu111 torchvision==0.10.1+cu111 torchaudio==0.9.1 -f https://download.pytorch.org/whl/torch_stable.html
2. Install pb_robot using the instructions below, not the ones in the repo's README
   1. Clone pb_robot outside of latent_feasibility.
   2. Compile the IKFast library for the panda
      1. cd pb_robot/src/pb_robot/ikfast/franka_panda
      2. python3 setup.py build (for errors, see troubleshooting below)
3. Install pddlstream
   1. Clone pddlstream outside of latent_feasibility.
   2. Follow installation instructions there.
4. Create a symlink to required repos (this assumes you cloned pb_robot and pddl_stream to your home directory). Run these commands from the top directory latent_feasibility.
   1. ln -s <path-to-pb_robot>/pb_robot/src/pb_robot .
   2. ln -s <path-to-pddstream>/pddlstream/pddlstream .

Grasping Dataset

To use the grasping modules (specifically from pb_robot/planners/antipodalGraspPlanner.py), you will need to download a dataset containing URDFs:

1. Download URDFs and object meshes from here.
2. Unzip the archive: unzip shapenet-sim.zip
3. Set the SHAPENET_ROOT environment variable to the location of the dataset: export SHAPENET_ROOT=<path-to-dataset>/shapenet-sem

Run

Troubleshooting

Update: This repo will now work by default with Python3.7. This troubleshooting may still be useful for setting up the repo using a different Python version.

On macOS Catalina using a Python3.7 virtualenv, building pb_robot with python setup.py build failed with the following error

./ikfast.h:41:10: fatal error: 'python3.6/Python.h' file not found

The compiler can't find the appropriate python header. The solution is to first locate the header:

$ find /usr/local/Cellar/ -name Python.h
/usr/local/Cellar//python/3.7.7/Frameworks/Python.framework/Versions/3.7/include/python3.7m/Python.h
/usr/local/Cellar//python@3.8/3.8.2/Frameworks/Python.framework/Versions/3.8/include/python3.8/Python.h

which prints the python include directories. I wanted to use 3.7, so then I set the environment variable

export CPLUS_INCLUDE_PATH=/usr/local/Cellar//python/3.7.7/Frameworks/Python.framework/Versions/3.7/include/

and finally modify pb_robot/src/pb_robot/ikfast/ikfast.h by changing

#include "python3.6/Python.h" -> #include "python3.7m/Python.h"