Skip to content

Instantly share code, notes, and snippets.

@taka-wang

taka-wang/3d_mesh.ipynb

Last active March 23, 2020 10:24
Show Gist options
  • Save taka-wang/bd9b22db366333a153a6d1340e8dda4c to your computer and use it in GitHub Desktop.
Save taka-wang/bd9b22db366333a153a6d1340e8dda4c to your computer and use it in GitHub Desktop.
Download ZIP
ai basic1 snippets
Display the source blob
Display the rendered blob
Raw
3d_mesh.ipynb
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
Display the source blob
Display the rendered blob
Raw
activation.ipynb
{
"cells": [
{
"cell_type": "code",
"execution_count": 4,
"metadata": {},
"outputs": [],
"source": [
"import matplotlib.pyplot as plt\n",
"import numpy as np"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Activation functions"
]
},
{
"cell_type": "code",
"execution_count": 5,
"metadata": {},
"outputs": [],
"source": [
"def tanh(x):\n",
" return (np.exp(x)-np.exp(-x))/(np.exp(x)+np.exp(-x))"
]
},
{
"cell_type": "code",
"execution_count": 6,
"metadata": {},
"outputs": [],
"source": [
"def tanhDerivative(x):\n",
" return 1.0 - tanh(x)**2"
]
},
{
"cell_type": "code",
"execution_count": 7,
"metadata": {},
"outputs": [],
"source": [
"def sigmoid(x):\n",
" return 1. / (1. + np.exp(-x))"
]
},
{
"cell_type": "code",
"execution_count": 8,
"metadata": {},
"outputs": [],
"source": [
"# Derivative of the Sigmoid activation function\n",
"def SigmoidDerivative(x):\n",
" return sigmoid(x) * (1 - sigmoid(x))"
]
},
{
"cell_type": "code",
"execution_count": 9,
"metadata": {},
"outputs": [],
"source": [
"def relu(x):\n",
" return np.maximum(0,x)"
]
},
{
"cell_type": "code",
"execution_count": 10,
"metadata": {},
"outputs": [],
"source": [
"# Derivative of the RELU activation function\n",
"def reluDerivative(x):\n",
" return np.where(x <= 0, 0, 1)"
]
},
{
"cell_type": "code",
"execution_count": 11,
"metadata": {},
"outputs": [],
"source": [
"def leaky(x):\n",
" return np.array([0.07*item if item<0 else item for item in x ]) "
]
},
{
"cell_type": "code",
"execution_count": 19,
"metadata": {},
"outputs": [
{
"data": {
"image/png": "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\n",
"text/plain": [
"<Figure size 432x288 with 1 Axes>"
]
},
"metadata": {
"needs_background": "dark"
},
"output_type": "display_data"
}
],
"source": [
"fig = plt.figure()\n",
"ax = fig.add_subplot(111)\n",
"\n",
"# generate data\n",
"x = np.arange(-6.1, 6.1, 0.1)\n",
"y = sigmoid(x)\n",
"ax.plot(x, y)\n",
"plt.xticks(np.arange(-6.0, 7.0, 1))\n",
"plt.yticks(np.arange(-0.2, 1.2, 0.2))\n",
"\n",
"# red line\n",
"plt.axvline(x=0, label='line at x = {}'.format(0), c='r')\n",
"\n",
"#''' change style\n",
"# ref: https://stackoverflow.com/questions/1982770/matplotlib-changing-the-color-of-an-axis\n",
"ax.spines['bottom'].set_color('#dddddd')\n",
"ax.spines['top'].set_color('#dddddd') \n",
"ax.spines['right'].set_color('#dddddd')\n",
"ax.spines['left'].set_color('#dddddd')\n",
"ax.tick_params(axis='x', colors='white')\n",
"ax.tick_params(axis='y', colors='white')\n",
"ax.yaxis.label.set_color('white')\n",
"ax.xaxis.label.set_color('white')\n",
"#'''\n",
"\n",
"# export\n",
"plt.grid()\n",
"plt.savefig('sigmoid.png', dpi=300, transparent=True)\n",
"plt.show()"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"fig = plt.figure()\n",
"ax = fig.add_subplot(111)\n",
"\n",
"# generate data\n",
"x = np.arange(-6.1, 6.1, 0.1)\n",
"#y = sigmoid(x)\n",
"y = SigmoidDerivative(x)\n",
"ax.plot(x, y)\n",
"plt.xticks(np.arange(-6.0, 7.0, 1))\n",
"plt.yticks(np.arange(-0.2, 1.2, 0.2))\n",
"\n",
"# red line\n",
"plt.axvline(x=0, label='line at x = {}'.format(0), c='r')\n",
"\n",
"#''' change style\n",
"# ref: https://stackoverflow.com/questions/1982770/matplotlib-changing-the-color-of-an-axis\n",
"ax.spines['bottom'].set_color('#dddddd')\n",
"ax.spines['top'].set_color('#dddddd') \n",
"ax.spines['right'].set_color('#dddddd')\n",
"ax.spines['left'].set_color('#dddddd')\n",
"ax.tick_params(axis='x', colors='white')\n",
"ax.tick_params(axis='y', colors='white')\n",
"ax.yaxis.label.set_color('white')\n",
"ax.xaxis.label.set_color('white')\n",
"#'''\n",
"\n",
"# export\n",
"plt.grid()\n",
"plt.savefig('derivative_sigmoid.png', dpi=300, transparent=True)\n",
"plt.show()"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"fig = plt.figure()\n",
"ax = fig.add_subplot(111)\n",
"\n",
"# generate data\n",
"x = np.arange(-6.1, 6.1, 0.1)\n",
"y = tanh(x)\n",
"ax.plot(x, y)\n",
"plt.xticks(np.arange(-6.0, 7.0, 1))\n",
"plt.yticks(np.arange(-1.2, 1.2, 0.4))\n",
"\n",
"# red line\n",
"plt.axvline(x=0, label='line at x = {}'.format(0), c='r')\n",
"\n",
"#''' change style\n",
"# ref: https://stackoverflow.com/questions/1982770/matplotlib-changing-the-color-of-an-axis\n",
"ax.spines['bottom'].set_color('#dddddd')\n",
"ax.spines['top'].set_color('#dddddd') \n",
"ax.spines['right'].set_color('#dddddd')\n",
"ax.spines['left'].set_color('#dddddd')\n",
"ax.tick_params(axis='x', colors='white')\n",
"ax.tick_params(axis='y', colors='white')\n",
"ax.yaxis.label.set_color('white')\n",
"ax.xaxis.label.set_color('white')\n",
"#'''\n",
"\n",
"# export\n",
"plt.grid()\n",
"plt.savefig('tanh.png', dpi=300, transparent=True)\n",
"plt.show()"
]
},
{
"cell_type": "code",
"execution_count": 18,
"metadata": {},
"outputs": [
{
"data": {
"image/png": "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\n",
"text/plain": [
"<Figure size 432x288 with 1 Axes>"
]
},
"metadata": {
"needs_background": "dark"
},
"output_type": "display_data"
}
],
"source": [
"fig = plt.figure()\n",
"ax = fig.add_subplot(111)\n",
"\n",
"# generate data\n",
"x = np.arange(-6.1, 6.1, 0.1)\n",
"y = tanhDerivative(x)\n",
"ax.plot(x, y)\n",
"plt.xticks(np.arange(-6.0, 7.0, 1))\n",
"plt.yticks(np.arange(0, 1.2, 0.2))\n",
"\n",
"# red line\n",
"plt.axvline(x=0, label='line at x = {}'.format(0), c='r')\n",
"\n",
"#''' change style\n",
"# ref: https://stackoverflow.com/questions/1982770/matplotlib-changing-the-color-of-an-axis\n",
"ax.spines['bottom'].set_color('#dddddd')\n",
"ax.spines['top'].set_color('#dddddd') \n",
"ax.spines['right'].set_color('#dddddd')\n",
"ax.spines['left'].set_color('#dddddd')\n",
"ax.tick_params(axis='x', colors='white')\n",
"ax.tick_params(axis='y', colors='white')\n",
"ax.yaxis.label.set_color('white')\n",
"ax.xaxis.label.set_color('white')\n",
"#'''\n",
"\n",
"# export\n",
"plt.grid()\n",
"plt.savefig('tanh_derivative.png', dpi=300, transparent=True)\n",
"plt.show()"
]
},
{
"cell_type": "code",
"execution_count": 14,
"metadata": {},
"outputs": [
{
"data": {
"image/png": "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\n",
"text/plain": [
"<Figure size 432x288 with 1 Axes>"
]
},
"metadata": {
"needs_background": "dark"
},
"output_type": "display_data"
}
],
"source": [
"fig = plt.figure()\n",
"ax = fig.add_subplot(111)\n",
"\n",
"# generate data\n",
"x = np.arange(-6.1, 6.1, 0.1)\n",
"y = relu(x)\n",
"ax.plot(x, y)\n",
"plt.xticks(np.arange(-6.0, 7.0, 1))\n",
"plt.yticks(np.arange(-2, 10, 2))\n",
"\n",
"# red line\n",
"plt.axvline(x=0, label='line at x = {}'.format(0), c='r')\n",
"\n",
"#''' change style\n",
"# ref: https://stackoverflow.com/questions/1982770/matplotlib-changing-the-color-of-an-axis\n",
"ax.spines['bottom'].set_color('#dddddd')\n",
"ax.spines['top'].set_color('#dddddd') \n",
"ax.spines['right'].set_color('#dddddd')\n",
"ax.spines['left'].set_color('#dddddd')\n",
"ax.tick_params(axis='x', colors='white')\n",
"ax.tick_params(axis='y', colors='white')\n",
"ax.yaxis.label.set_color('white')\n",
"ax.xaxis.label.set_color('white')\n",
"#'''\n",
"\n",
"# export\n",
"plt.grid()\n",
"plt.savefig('relu.png', dpi=300, transparent=True)\n",
"plt.show()"
]
},
{
"cell_type": "code",
"execution_count": 15,
"metadata": {},
"outputs": [
{
"data": {
"image/png": "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\n",
"text/plain": [
"<Figure size 432x288 with 1 Axes>"
]
},
"metadata": {
"needs_background": "dark"
},
"output_type": "display_data"
}
],
"source": [
"fig = plt.figure()\n",
"ax = fig.add_subplot(111)\n",
"\n",
"# generate data\n",
"x = np.arange(-6.1, 6.1, 0.1)\n",
"y = reluDerivative(x)\n",
"ax.plot(x, y)\n",
"plt.xticks(np.arange(-6.0, 7.0, 1))\n",
"plt.yticks(np.arange(0, 1.0, 0.2))\n",
"\n",
"# red line\n",
"plt.axvline(x=0, label='line at x = {}'.format(0), c='r')\n",
"\n",
"#''' change style\n",
"# ref: https://stackoverflow.com/questions/1982770/matplotlib-changing-the-color-of-an-axis\n",
"ax.spines['bottom'].set_color('#dddddd')\n",
"ax.spines['top'].set_color('#dddddd') \n",
"ax.spines['right'].set_color('#dddddd')\n",
"ax.spines['left'].set_color('#dddddd')\n",
"ax.tick_params(axis='x', colors='white')\n",
"ax.tick_params(axis='y', colors='white')\n",
"ax.yaxis.label.set_color('white')\n",
"ax.xaxis.label.set_color('white')\n",
"#'''\n",
"\n",
"# export\n",
"plt.grid()\n",
"plt.savefig('relu_derivative.png', dpi=300, transparent=True)\n",
"plt.show()"
]
},
{
"cell_type": "code",
"execution_count": 16,
"metadata": {},
"outputs": [
{
"data": {
"image/png": "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\n",
"text/plain": [
"<Figure size 432x288 with 1 Axes>"
]
},
"metadata": {
"needs_background": "dark"
},
"output_type": "display_data"
}
],
"source": [
"fig = plt.figure()\n",
"ax = fig.add_subplot(111)\n",
"\n",
"# generate data\n",
"x = np.arange(-6.1, 6.1, 0.1)\n",
"y = leaky(x)\n",
"ax.plot(x, y)\n",
"plt.xticks(np.arange(-6.0, 7.0, 1))\n",
"plt.yticks(np.arange(-2, 10, 2))\n",
"\n",
"# red line\n",
"plt.axvline(x=0, label='line at x = {}'.format(0), c='r')\n",
"\n",
"#''' change style\n",
"# ref: https://stackoverflow.com/questions/1982770/matplotlib-changing-the-color-of-an-axis\n",
"ax.spines['bottom'].set_color('#dddddd')\n",
"ax.spines['top'].set_color('#dddddd') \n",
"ax.spines['right'].set_color('#dddddd')\n",
"ax.spines['left'].set_color('#dddddd')\n",
"ax.tick_params(axis='x', colors='white')\n",
"ax.tick_params(axis='y', colors='white')\n",
"ax.yaxis.label.set_color('white')\n",
"ax.xaxis.label.set_color('white')\n",
"#'''\n",
"\n",
"# export\n",
"plt.grid()\n",
"plt.savefig('leaky.png', dpi=300, transparent=True)\n",
"plt.show()"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": []
}
],
"metadata": {
"kernelspec": {
"display_name": "Python 3",
"language": "python",
"name": "python3"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.7.6"
}
},
"nbformat": 4,
"nbformat_minor": 4
}
Display the source blob
Display the rendered blob
Raw
axis.ipynb
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
Display the source blob
Display the rendered blob
Raw
broadcasting.ipynb
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
Display the source blob
Display the rendered blob
Raw
entropy.ipynb
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
Display the source blob
Display the rendered blob
Raw
mnist.ipynb
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
Display the source blob
Display the rendered blob
Raw
numpy_speed.ipynb
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
Display the source blob
Display the rendered blob
Raw
torch_speed.ipynb
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment