Navigation Menu

Skip to content

Instantly share code, notes, and snippets.

@yang-zhang

yang-zhang/08_pets_tutorial-yz-20190911-dbg-image-shape.ipynb

Last active September 11, 2019 16:21
Show Gist options
  • Star 0 You must be signed in to star a gist
  • Fork 0 You must be signed in to fork a gist
  • Save yang-zhang/bea20a7e9791c341507e934e53a107fe to your computer and use it in GitHub Desktop.
Save yang-zhang/bea20a7e9791c341507e934e53a107fe to your computer and use it in GitHub Desktop.
Download ZIP
Debugging fastai_dev 08_pets_tutorial_image_dim
Display the source blob
Display the rendered blob
Raw
08_pets_tutorial-yz-20190911-dbg-image-shape.ipynb
{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"https://github.com/fastai/fastai_dev/blob/master/dev/08_pets_tutorial.ipynb"
]
},
{
"cell_type": "code",
"execution_count": 1,
"metadata": {},
"outputs": [],
"source": [
"from local.torch_basics import *\n",
"from local.test import *\n",
"from local.data.all import *\n",
"from local.vision.core import *\n",
"from local.notebook.showdoc import *"
]
},
{
"cell_type": "code",
"execution_count": 2,
"metadata": {
"scrolled": false
},
"outputs": [
{
"data": {
"text/plain": [
"'https://s3.amazonaws.com/fast-ai-imageclas/oxford-iiit-pet.tgz'"
]
},
"execution_count": 2,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"URLs.PETS"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"get image files and count dimension; most images are 3d but some are 2d."
]
},
{
"cell_type": "code",
"execution_count": 3,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"Counter({2: 9, 3: 7381})"
]
},
"execution_count": 3,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"source = untar_data(URLs.PETS)/\"images\"\n",
"items = get_image_files(source)\n",
"item2ndim = {o:array(Image.open(o)).ndim for o in items}\n",
"\n",
"Counter(item2ndim.values())"
]
},
{
"cell_type": "code",
"execution_count": 10,
"metadata": {},
"outputs": [],
"source": [
"items_2d = [k for k,v in item2ndim.items() if v==2]\n",
"items_3d = list(set(item2ndim.keys())-set(items_2d))"
]
},
{
"cell_type": "code",
"execution_count": 11,
"metadata": {
"scrolled": false
},
"outputs": [
{
"data": {
"text/plain": [
"array([[ 62, 44, 62, ..., 105, 105, 105],\n",
" [ 62, 62, 44, ..., 60, 60, 60],\n",
" [ 44, 62, 44, ..., 60, 196, 105],\n",
" ...,\n",
" [183, 223, 206, ..., 76, 181, 76],\n",
" [ 84, 209, 224, ..., 136, 76, 181],\n",
" [223, 183, 223, ..., 76, 76, 136]], dtype=uint8)"
]
},
"execution_count": 11,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"array(Image.open(items_2d[0]))"
]
},
{
"cell_type": "code",
"execution_count": 12,
"metadata": {
"collapsed": true
},
"outputs": [
{
"data": {
"text/plain": [
"array([[[ 14, 4, 3],\n",
" [ 14, 4, 3],\n",
" [ 14, 4, 3],\n",
" ...,\n",
" [ 42, 29, 12],\n",
" [ 42, 29, 12],\n",
" [ 42, 29, 12]],\n",
"\n",
" [[ 14, 4, 3],\n",
" [ 14, 4, 3],\n",
" [ 14, 4, 3],\n",
" ...,\n",
" [ 42, 29, 12],\n",
" [ 42, 29, 12],\n",
" [ 42, 29, 12]],\n",
"\n",
" [[ 14, 4, 3],\n",
" [ 14, 4, 3],\n",
" [ 14, 4, 3],\n",
" ...,\n",
" [ 42, 29, 12],\n",
" [ 42, 29, 12],\n",
" [ 42, 29, 12]],\n",
"\n",
" ...,\n",
"\n",
" [[234, 226, 224],\n",
" [226, 218, 216],\n",
" [220, 212, 210],\n",
" ...,\n",
" [183, 175, 162],\n",
" [180, 176, 164],\n",
" [181, 179, 166]],\n",
"\n",
" [[239, 231, 229],\n",
" [236, 228, 226],\n",
" [234, 226, 224],\n",
" ...,\n",
" [166, 158, 145],\n",
" [182, 178, 166],\n",
" [199, 197, 184]],\n",
"\n",
" [[234, 226, 224],\n",
" [235, 227, 225],\n",
" [237, 229, 227],\n",
" ...,\n",
" [196, 188, 175],\n",
" [211, 207, 195],\n",
" [218, 216, 203]]], dtype=uint8)"
]
},
"execution_count": 12,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"array(Image.open(items_3d[0]))"
]
},
{
"cell_type": "code",
"execution_count": 13,
"metadata": {},
"outputs": [],
"source": [
"def resized_image(fn:Path, sz=128):\n",
" x = Image.open(fn).resize((sz,sz))\n",
" # Convert image to tensor for modeling\n",
" return tensor(array(x)).permute(2,0,1).float()/255."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"`resized_image` works fine for 3d arrays"
]
},
{
"cell_type": "code",
"execution_count": 8,
"metadata": {},
"outputs": [],
"source": [
"_ = [resized_image(o) for o in items_3d]"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"but error on 2d arrays"
]
},
{
"cell_type": "code",
"execution_count": 14,
"metadata": {},
"outputs": [
{
"ename": "RuntimeError",
"evalue": "number of dims don't match in permute",
"output_type": "error",
"traceback": [
"\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
"\u001b[0;31mRuntimeError\u001b[0m Traceback (most recent call last)",
"\u001b[0;32m<ipython-input-14-fb519d26debf>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[0;32m----> 1\u001b[0;31m \u001b[0mresized_image\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mitems_2d\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;36m0\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m",
"\u001b[0;32m<ipython-input-13-4b2ef1648cbb>\u001b[0m in \u001b[0;36mresized_image\u001b[0;34m(fn, sz)\u001b[0m\n\u001b[1;32m 2\u001b[0m \u001b[0mx\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mImage\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mopen\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mfn\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mresize\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0msz\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0msz\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 3\u001b[0m \u001b[0;31m# Convert image to tensor for modeling\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m----> 4\u001b[0;31m \u001b[0;32mreturn\u001b[0m \u001b[0mtensor\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0marray\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mx\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mpermute\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;36m2\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;36m0\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;36m1\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mfloat\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m/\u001b[0m\u001b[0;36m255.\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m",
"\u001b[0;31mRuntimeError\u001b[0m: number of dims don't match in permute"
]
}
],
"source": [
"resized_image(items_2d[0]) "
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"list of 2d images"
]
},
{
"cell_type": "code",
"execution_count": 16,
"metadata": {
"scrolled": true
},
"outputs": [
{
"data": {
"text/plain": [
"['Egyptian_Mau_167',\n",
" 'Egyptian_Mau_177',\n",
" 'Egyptian_Mau_139',\n",
" 'Egyptian_Mau_129',\n",
" 'Abyssinian_34',\n",
" 'Egyptian_Mau_191',\n",
" 'staffordshire_bull_terrier_2',\n",
" 'Egyptian_Mau_145',\n",
" 'staffordshire_bull_terrier_22']"
]
},
"execution_count": 16,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"[o.stem for o in items_2d]"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": []
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": []
}
],
"metadata": {
"kernelspec": {
"display_name": "fastai_dev",
"language": "python",
"name": "fastai_dev"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.7.4"
},
"toc": {
"base_numbering": 1,
"nav_menu": {},
"number_sections": true,
"sideBar": true,
"skip_h1_title": false,
"title_cell": "Table of Contents",
"title_sidebar": "Contents",
"toc_cell": false,
"toc_position": {},
"toc_section_display": true,
"toc_window_display": true
}
},
"nbformat": 4,
"nbformat_minor": 2
}
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment