nafeesb

Images and Matrices

Load image with imageio

import imageio
import matplotlib.pyplot as plt
img = imageio.imread('file.jpg')
plt.imshow(img)

Load image with PIL

nafeesb

# good for testing frameworks

import maya.standalone
maya.standalone.initialize()
 
from maya import cmds
cmds.loadPlugin('path/to/plugin.so')

nafeesb

#Uploading the Dataset 
from google.colab import files
uploaded = files.upload()

#Save uploaded file on the Virtual Machine's 
#Thanks to user3800642 from StackOverflow 

with open("breast_cancer.csv", 'w') as f:
    f.write(uploaded[uploaded.keys()[0]])

nafeesb

# Restart the VM
!kill -9 -1

## Mount a google drive
# install prereqs
!apt-get install -y -qq software-properties-common python-software-properties module-init-tools
!add-apt-repository -y ppa:alessandro-strada/ppa 2>&1 > /dev/null
!apt-get update -qq 2>&1 > /dev/null
!apt-get -y install -qq google-drive-ocamlfuse fuse
from google.colab import auth

nafeesb

///
/// Average max distance
///
float maxdistance = 15;
int maxpoints = 5;
int closept[] = pcfind_radius(0, "P", "pscale", 1.0, @P, maxdistance, maxpoints);

f@dist = 0;
foreach (int opt; closept) {
    vector npos = point(0, "P", opt);

nafeesb

CUDA

Free GPU memory

del object_using_gpu
torch.cuda.empty_cache()

Tensor memory use

nafeesb

    void branchlessONB( Vector3 N, ref Vector3 A, ref Vector3 B)
    {
        float sign = N.z >= 0.0f ? 1.0f : -1.0f;
        float a = 1.0f / (sign + N.z);
        float b = N.x * N.y * a;
        A.Set(1.0f + sign * N.x * N.x * a, sign * b, -sign * N.x);
        B.Set(b, sign + N.y * N.y * a, -N.y);
    }
    

nafeesb

import numpy as np
import torch
from torch import nn
from itertools import count
import time

#%%
# make some data
N = 100

nafeesb

Anaconda Configuration

conda create -n tensorflow tensorflow-gpu pip python=3.6
conda install ipython matplotlib cython mkl_random mkl mkl-devel swig

Install OpenAI Gym

Install from source: https://github.com/openai/gym

pip install -e .

nafeesb

Ouput Zero for divide by zero

length_err_pct = np.divide( pred_length, true_length, out=np.zeros_like(pred_length), where=true_length!=0)

Random subset of indices

sel = np.random.choice(testX.shape[0], samples, replace=False)

Save formatting