Skip to content

Instantly share code, notes, and snippets.

@oguiza
Created November 21, 2018 10:45
Show Gist options
  • Save oguiza/3d8266e04767b119f7c13ccadd071862 to your computer and use it in GitHub Desktop.
Save oguiza/3d8266e04767b119f7c13ccadd071862 to your computer and use it in GitHub Desktop.
course-v3/my_nbs/dl1/Untitled.ipynb
Display the source blob
Display the rendered blob
Raw
{
"cells": [
{
"metadata": {},
"cell_type": "markdown",
"source": "# TSSG Fast Learning Competition: Earthquakes*"
},
{
"metadata": {
"ExecuteTime": {
"start_time": "2018-11-21T08:25:16.022459Z",
"end_time": "2018-11-21T08:25:16.026724Z"
}
},
"cell_type": "markdown",
"source": "*A. Bagnall, J. Lines, W. Vickers and E. Keogh, The UEA & UCR Time Series Classification Repository,\nwww.timeseriesclassification.com"
},
{
"metadata": {
"ExecuteTime": {
"start_time": "2018-11-21T10:25:37.578229Z",
"end_time": "2018-11-21T10:25:37.792549Z"
},
"trusted": true
},
"cell_type": "code",
"source": "%reload_ext autoreload\n%autoreload 2\n%matplotlib inline",
"execution_count": 1,
"outputs": []
},
{
"metadata": {
"ExecuteTime": {
"start_time": "2018-11-21T10:25:38.505040Z",
"end_time": "2018-11-21T10:25:39.606206Z"
},
"trusted": true
},
"cell_type": "code",
"source": "from fastai import *\nimport fastai\nfastai.__version__",
"execution_count": 2,
"outputs": [
{
"output_type": "execute_result",
"execution_count": 2,
"data": {
"text/plain": "'1.0.28'"
},
"metadata": {}
}
]
},
{
"metadata": {
"ExecuteTime": {
"start_time": "2018-11-21T10:25:41.681343Z",
"end_time": "2018-11-21T10:25:41.735200Z"
},
"trusted": true
},
"cell_type": "code",
"source": "import zipfile\nimport tempfile\nimport shutil\nimport os\nimport sys\nimport csv\ntry:\n from urllib import urlretrieve\nexcept ImportError:\n from urllib.request import urlretrieve\ntry:\n from zipfile import BadZipFile as BadZipFile\nexcept ImportError:\n from zipfile import BadZipfile as BadZipFile\n\nfrom tslearn.utils import to_time_series_dataset",
"execution_count": 3,
"outputs": []
},
{
"metadata": {
"ExecuteTime": {
"start_time": "2018-11-21T10:25:42.833509Z",
"end_time": "2018-11-21T10:25:42.862437Z"
},
"trusted": true
},
"cell_type": "code",
"source": "def extract_from_zip_url(url, target_dir=None, verbose=False):\n \"\"\"Download a zip file from its URL and unzip it.\"\"\"\n fname = os.path.basename(url)\n tmpdir = tempfile.mkdtemp()\n local_zip_fname = os.path.join(tmpdir, fname)\n urlretrieve(url, local_zip_fname)\n try:\n if not os.path.exists(target_dir):\n os.makedirs(target_dir)\n zipfile.ZipFile(local_zip_fname, \"r\").extractall(path=target_dir)\n shutil.rmtree(tmpdir)\n if verbose:\n print(\"Successfully extracted file %s to path %s\" % (local_zip_fname, target_dir))\n return target_dir\n except BadZipFile:\n shutil.rmtree(tmpdir)\n if verbose:\n sys.stderr.write(\"Corrupted zip file encountered, aborting.\\n\")\n return None",
"execution_count": 4,
"outputs": []
},
{
"metadata": {
"ExecuteTime": {
"start_time": "2018-11-21T10:25:44.161148Z",
"end_time": "2018-11-21T10:25:44.191811Z"
},
"trusted": true
},
"cell_type": "code",
"source": "def prepare_dataset(dataset_name, target_dir):\n '''\n Download selected UCR dataset, unzip files to target dir, \n read train & test files, and normalizes data\n '''\n\n full_path = os.path.join(target_dir, dataset_name)\n fname_train = dataset_name + \"_TRAIN.txt\"\n fname_test = dataset_name + \"_TEST.txt\"\n if not os.path.exists(os.path.join(full_path, fname_train)) or \\\n not os.path.exists(os.path.join(full_path, fname_test)):\n url = \"http://www.timeseriesclassification.com/Downloads/%s.zip\" % dataset_name\n for fname in [fname_train, fname_test]:\n if os.path.exists(os.path.join(full_path, fname)):\n os.remove(os.path.join(full_path, fname))\n extract_from_zip_url(url, target_dir=full_path, verbose=False)\n try:\n data_train = np.loadtxt(os.path.join(full_path, fname_train), delimiter=None)\n data_test = np.loadtxt(os.path.join(full_path, fname_test), delimiter=None)\n except:\n return None, None, None, None\n\n X_train = to_time_series_dataset(data_train[:, 1:])\n y_train = data_train[:, 0].astype(np.int)\n X_test = to_time_series_dataset(data_test[:, 1:])\n y_test = data_test[:, 0].astype(np.int)\n\n X_train = np.squeeze(X_train)\n # scale the values\n X_train_mean = np.mean(X_train)\n X_train_std = np.std(X_train)\n X_train = (X_train - X_train_mean) / X_train_std\n\n nb_classes = len(np.unique(y_train))\n y_train = ((y_train - y_train.min()) / (y_train.max() - y_train.min()) * (nb_classes - 1)).astype(int)\n y_train = y_train.ravel()\n\n X_test = np.squeeze(X_test)\n # scale the values\n X_test = (X_test - X_train_mean) / X_train_std\n\n y_test = ((y_test - y_test.min()) / (y_test.max() - y_test.min()) * (nb_classes - 1)).astype(int)\n y_test = y_test.ravel()\n\n return X_train, y_train, X_test, y_test",
"execution_count": 5,
"outputs": []
},
{
"metadata": {},
"cell_type": "markdown",
"source": "## Prepare data"
},
{
"metadata": {
"ExecuteTime": {
"start_time": "2018-11-21T10:25:45.849084Z",
"end_time": "2018-11-21T10:25:45.869899Z"
},
"trusted": true
},
"cell_type": "code",
"source": "SEL_DATASET = 'Earthquakes'\nTGT_DIR = Path('my_data/UCR_univariate/' + SEL_DATASET)",
"execution_count": 6,
"outputs": []
},
{
"metadata": {},
"cell_type": "markdown",
"source": "prepare_dataset() will download url to selected dataset, unzip file, load them to sel target dir, prepare train & test sets and normalize data."
},
{
"metadata": {
"ExecuteTime": {
"start_time": "2018-11-21T10:25:47.304723Z",
"end_time": "2018-11-21T10:25:47.470052Z"
},
"trusted": true
},
"cell_type": "code",
"source": "X_train, y_train, X_test, y_test = prepare_dataset(SEL_DATASET, TGT_DIR)",
"execution_count": 7,
"outputs": []
},
{
"metadata": {
"ExecuteTime": {
"start_time": "2018-11-21T10:25:48.680753Z",
"end_time": "2018-11-21T10:25:48.703020Z"
},
"trusted": true
},
"cell_type": "code",
"source": "print(\"Number of train samples: \", X_train.shape[0], \"Number of test samples: \", X_test.shape[0])\nprint(\"Sequence length: \", X_train.shape[-1])",
"execution_count": 8,
"outputs": [
{
"output_type": "stream",
"text": "Number of train samples: 322 Number of test samples: 139\nSequence length: 512\n",
"name": "stdout"
}
]
},
{
"metadata": {
"ExecuteTime": {
"start_time": "2018-11-21T10:39:42.879214Z",
"end_time": "2018-11-21T10:39:43.139137Z"
},
"trusted": true
},
"cell_type": "code",
"source": "f, (ax1, ax2) = plt.subplots(1, 2, sharey=True, figsize=(15,5))\nax1.plot(X_train[1])\nax1.set_title('class ' + str(y_train[1]))\nax2.plot(X_train[0])\nax2.set_title('class ' + str(y_train[0]))\nplt.show()",
"execution_count": 38,
"outputs": [
{
"output_type": "display_data",
"data": {
"text/plain": "<Figure size 1080x360 with 2 Axes>",
"image/png": "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\n"
},
"metadata": {
"needs_background": "light"
}
}
]
},
{
"metadata": {},
"cell_type": "markdown",
"source": "You are now ready to start creating working on this dataset!!\n\nAccording to the UCR website, the current state of the art accuracy on this dataset is 0.835526316, achieved with a non deep learning model ('HIVE-COTE'). Will you be able to improve that?"
},
{
"metadata": {
"trusted": true
},
"cell_type": "code",
"source": "",
"execution_count": null,
"outputs": []
}
],
"metadata": {
"kernelspec": {
"name": "fastai-v1",
"display_name": "fastai-v1",
"language": "python"
},
"toc": {
"nav_menu": {},
"number_sections": true,
"sideBar": true,
"skip_h1_title": false,
"base_numbering": 1,
"title_cell": "Table of Contents",
"title_sidebar": "Contents",
"toc_cell": false,
"toc_position": {},
"toc_section_display": true,
"toc_window_display": false
},
"language_info": {
"name": "python",
"version": "3.7.0",
"mimetype": "text/x-python",
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"pygments_lexer": "ipython3",
"nbconvert_exporter": "python",
"file_extension": ".py"
},
"gist": {
"id": "",
"data": {
"description": "course-v3/my_nbs/dl1/Untitled.ipynb",
"public": true
}
}
},
"nbformat": 4,
"nbformat_minor": 2
}
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment