DanielTakeshi

Guide to Reporting Errors (CSCI 545, Spring 2024)

Note: this guide has been written for CSCI 545, Introduction to Robotics, taught by Daniel Seita in Spring 2024. We have a policy that students must state that they have read this guide before requesting software help. This is designed for the benefit of both the course staff but also for the students.

This is a short guide on how to report code-related errors. If you run into errors, following these guidelines will increase the chances that your issue gets quickly resolved. Otherwise, it might result in a protracted

DanielTakeshi

"""
Copyright (c) 2020, NVIDIA CORPORATION. All rights reserved.

NVIDIA CORPORATION and its licensors retain all intellectual property
and proprietary rights in and to this software, related documentation
and any modifications thereto. Any use, reproduction, disclosure or
distribution of this software and related documentation without an express
license agreement from NVIDIA CORPORATION is strictly prohibited.

Franka Cube Pick

DanielTakeshi

"""
Runs IK to get the UR5 end-effector to reach a target. Inspect CPU vs GPU mode.
Use a yellow sphere to show the tip of the end-effector, and a blue sphere to
show the target. These spheres are only used for debugging / visualization.
"""
from isaacgym import gymapi
from isaacgym import gymutil
from isaacgym import gymtorch
from isaacgym.torch_utils import (quat_conjugate, quat_mul, quat_apply)
import numpy as np

DanielTakeshi

import cv2
import rospy
import argparse
import numpy as np
import matplotlib.pyplot as plt
from cv_bridge import CvBridge
from sensor_msgs.msg import Image

class Segmenter():
    def __init__(self, args):

DanielTakeshi

import matplotlib
matplotlib.use('Agg')
import matplotlib.pyplot as plt
plt.style.use('seaborn')
import numpy as np

nrows, ncols = 1, 2
fig, ax = plt.subplots(nrows, ncols, sharey=False, squeeze=True, figsize=(9*ncols, 6*nrows))
x = np.arange(100)
y0 = x + np.random.normal(loc=0.0, scale=10.0, size=100)

DanielTakeshi

import gym
import numpy as np
import imageio

# Make the environment and the initial observation `o`.
env_name = 'Walker2d-v3'
env = gym.make(env_name)
obs_dim = env.observation_space.shape
act_dim = env.action_space.shape[0]
act_limit = env.action_space.high[0]

DanielTakeshi

"""
How we might sample from a log-uniform distribution
https://stats.stackexchange.com/questions/155552/what-does-log-uniformly-distribution-mean

Only run one of these three cases at a time, otherwise the plots update each
other. Run with these versions:

matplotlib      3.2.1
numpy           1.18.3
"""

DanielTakeshi

"""Some exception testing.

The code in the `try` block will stop as soon as any exception is encountered.

https://realpython.com/python-exceptions/
https://realpython.com/the-most-diabolical-python-antipattern/
"""

a = 5

DanielTakeshi

I am testing different PyTorch transforms on my data, and saving sample
images as they are loaded during training. See examples below.

DanielTakeshi

import torch
import torch.nn as nn
import torch.optim as optim
from torch.optim import lr_scheduler
import torchvision.models as models
from torchvision import datasets, transforms
import copy, cv2, os, sys, pickle, time
import numpy as np
from os.path import join