![\"Open](\"https://colab.research.google.com/assets/colab-badge.svg\"){kind=icon}
"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "Et58FCcZx8uL"
+ },
+ "source": [
+ "# Location Heatmaps under Distributed Differential Privacy with Secure Aggregation\n",
+ "\n",
+ "This code shows examples of the proposed algorithm from [PETS'22 paper](https://arxiv.org/abs/2111.02356). We use publicly available [map](https://www.nytimes.com/2018/12/14/reader-center/phone-data-location-investigation.html) from the NYTimes article and population density from [Facebook Humanitarian Data Project](https://data.humdata.org/organization/facebook).\n",
+ "\n",
+ "\n",
+ "To cite our work use:\n",
+ "```\n",
+ "@inproceedings{bagdasaryan2022sparse,\n",
+ " title={Towards Sparse Federated Analytics: Location Heatmaps under Distributed Differential Privacy with Secure Aggregation}, \n",
+ " author={Eugene Bagdasaryan and Peter Kairouz and Stefan Mellem and Adrià Gascón and Kallista Bonawitz and Deborah Estrin and Marco Gruteser},\n",
+ " year={2022},\n",
+ " bootkitle={PETS},\n",
+ "}\n",
+ "```\n",
+ "\n",
+ "\n"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "colab": {
+ "base_uri": "https://localhost:8080/"
+ },
+ "id": "XjqIk04DMLl_",
+ "outputId": "5d6da059-6d8b-4282-ac54-ecee7a248daa"
+ },
+ "source": [
+ "!git clone https://github.com/google-research/federated.git"
+ ],
+ "execution_count": 1,
+ "outputs": [
+ {
+ "output_type": "stream",
+ "name": "stdout",
+ "text": [
+ "Cloning into 'federated'...\n",
+ "remote: Enumerating objects: 3701, done.\u001b[K\n",
+ "remote: Counting objects: 100% (1030/1030), done.\u001b[K\n",
+ "remote: Compressing objects: 100% (158/158), done.\u001b[K\n",
+ "remote: Total 3701 (delta 919), reused 883 (delta 872), pack-reused 2671\u001b[K\n",
+ "Receiving objects: 100% (3701/3701), 13.15 MiB | 15.90 MiB/s, done.\n",
+ "Resolving deltas: 100% (2595/2595), done.\n"
+ ]
+ }
+ ]
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "Bbfz7uFB-HCt"
+ },
+ "source": [
+ "%load_ext autoreload\n",
+ "%autoreload 2"
+ ],
+ "execution_count": 12,
+ "outputs": []
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "colab": {
+ "base_uri": "https://localhost:8080/"
+ },
+ "id": "LdD1HVo2NpTB",
+ "outputId": "60c117e6-600c-4aea-9331-470d35f720b1"
+ },
+ "source": [
+ "!pip install torch\n",
+ "!pip install matplotlib\n",
+ "!pip install tqdm\n",
+ "!pip install pygtrie"
+ ],
+ "execution_count": 13,
+ "outputs": [
+ {
+ "output_type": "stream",
+ "name": "stdout",
+ "text": [
+ "Looking in indexes: https://pypi.org/simple, https://us-python.pkg.dev/colab-wheels/public/simple/\n",
+ "Requirement already satisfied: torch in /usr/local/lib/python3.7/dist-packages (1.12.1+cu113)\n",
+ "Requirement already satisfied: typing-extensions in /usr/local/lib/python3.7/dist-packages (from torch) (4.1.1)\n",
+ "Looking in indexes: https://pypi.org/simple, https://us-python.pkg.dev/colab-wheels/public/simple/\n",
+ "Requirement already satisfied: matplotlib in /usr/local/lib/python3.7/dist-packages (3.2.2)\n",
+ "Requirement already satisfied: cycler>=0.10 in /usr/local/lib/python3.7/dist-packages (from matplotlib) (0.11.0)\n",
+ "Requirement already satisfied: python-dateutil>=2.1 in /usr/local/lib/python3.7/dist-packages (from matplotlib) (2.8.2)\n",
+ "Requirement already satisfied: pyparsing!=2.0.4,!=2.1.2,!=2.1.6,>=2.0.1 in /usr/local/lib/python3.7/dist-packages (from matplotlib) (3.0.9)\n",
+ "Requirement already satisfied: numpy>=1.11 in /usr/local/lib/python3.7/dist-packages (from matplotlib) (1.21.6)\n",
+ "Requirement already satisfied: kiwisolver>=1.0.1 in /usr/local/lib/python3.7/dist-packages (from matplotlib) (1.4.4)\n",
+ "Requirement already satisfied: typing-extensions in /usr/local/lib/python3.7/dist-packages (from kiwisolver>=1.0.1->matplotlib) (4.1.1)\n",
+ "Requirement already satisfied: six>=1.5 in /usr/local/lib/python3.7/dist-packages (from python-dateutil>=2.1->matplotlib) (1.15.0)\n",
+ "Looking in indexes: https://pypi.org/simple, https://us-python.pkg.dev/colab-wheels/public/simple/\n",
+ "Requirement already satisfied: tqdm in /usr/local/lib/python3.7/dist-packages (4.64.0)\n",
+ "Looking in indexes: https://pypi.org/simple, https://us-python.pkg.dev/colab-wheels/public/simple/\n",
+ "Collecting pygtrie\n",
+ " Downloading pygtrie-2.5.0.tar.gz (39 kB)\n",
+ "Building wheels for collected packages: pygtrie\n",
+ " Building wheel for pygtrie (setup.py) ... \u001b[?25l\u001b[?25hdone\n",
+ " Created wheel for pygtrie: filename=pygtrie-2.5.0-py3-none-any.whl size=20944 sha256=e82980bbb446b127e5c0822c748f7b4f739134a90e5329d62e5945ddc12d6154\n",
+ " Stored in directory: /root/.cache/pip/wheels/d1/5b/c8/7791f519f0b9f7186bd7531fe8749a849934739069ee2f2a43\n",
+ "Successfully built pygtrie\n",
+ "Installing collected packages: pygtrie\n",
+ "Successfully installed pygtrie-2.5.0\n"
+ ]
+ }
+ ]
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "mEWtFrMlM7sN"
+ },
+ "source": [
+ "import matplotlib.pyplot as plt\n",
+ "import torch\n",
+ "from tqdm.notebook import trange, tqdm"
+ ],
+ "execution_count": 14,
+ "outputs": []
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "jZ5lGxloNLvN"
+ },
+ "source": [
+ "import sys\n",
+ "sys.path.insert(1, \"federated/analytics/location_heatmaps\")"
+ ],
+ "execution_count": 15,
+ "outputs": []
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "APFZNJ9XOOlk"
+ },
+ "source": [
+ "from run_experiment import *\n",
+ "import matplotlib.pyplot as plt\n",
+ "import torch\n",
+ "from tqdm.notebook import trange, tqdm\n",
+ "import geo_utils\n",
+ "import metrics\n",
+ "import mechanisms"
+ ],
+ "execution_count": 16,
+ "outputs": []
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "colab": {
+ "base_uri": "https://localhost:8080/"
+ },
+ "id": "dQnTjDhXOOip",
+ "outputId": "87b71493-1a92-4a46-da52-9071c7c29a97"
+ },
+ "source": [
+ "# Original source:\n",
+ "# !wget https://static01.nyt.com/images/2018/12/14/business/10location-insider/10location-promo-superJumbo-v2.jpg\n",
+ "\n",
+ "!wget https://www.cs.cornell.edu/~eugene/assets/projects/location_heatmaps/location_heatmap.jpg"
+ ],
+ "execution_count": 17,
+ "outputs": [
+ {
+ "output_type": "stream",
+ "name": "stdout",
+ "text": [
+ "--2022-08-22 23:11:08-- https://www.cs.cornell.edu/~eugene/assets/projects/location_heatmaps/location_heatmap.jpg\n",
+ "Resolving www.cs.cornell.edu (www.cs.cornell.edu)... 132.236.207.36\n",
+ "Connecting to www.cs.cornell.edu (www.cs.cornell.edu)|132.236.207.36|:443... connected.\n",
+ "HTTP request sent, awaiting response... 200 OK\n",
+ "Length: 4924651 (4.7M) [image/jpeg]\n",
+ "Saving to: ‘location_heatmap.jpg’\n",
+ "\n",
+ "location_heatmap.jp 100%[===================>] 4.70M 2.50MB/s in 1.9s \n",
+ "\n",
+ "2022-08-22 23:11:12 (2.50 MB/s) - ‘location_heatmap.jpg’ saved [4924651/4924651]\n",
+ "\n"
+ ]
+ }
+ ]
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "id": "FgkiCaKbOOf2",
+ "outputId": "339d3b2f-f842-49f3-b21a-81cae50f3ec7",
+ "colab": {
+ "base_uri": "https://localhost:8080/"
+ }
+ },
+ "source": [
+ "true_image, dataset = get_data('location_heatmap.jpg')"
+ ],
+ "execution_count": 18,
+ "outputs": [
+ {
+ "output_type": "stream",
+ "name": "stderr",
+ "text": [
+ "100%|██████████| 1024/1024 [00:25<00:00, 40.96it/s]\n"
+ ]
+ }
+ ]
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "colab": {
+ "base_uri": "https://localhost:8080/"
+ },
+ "id": "pFTYc19MzGvo",
+ "outputId": "a41adce9-3291-4f95-f4c3-0c81b80700d8"
+ },
+ "source": [
+ "print(f'Total users on the map: {true_image.sum():,}.')"
+ ],
+ "execution_count": 19,
+ "outputs": [
+ {
+ "output_type": "stream",
+ "name": "stdout",
+ "text": [
+ "Total users on the map: 54,599,988.\n"
+ ]
+ }
+ ]
+ },
+ {
+ "cell_type": "code",
+ "metadata": {
+ "colab": {
+ "base_uri": "https://localhost:8080/",
+ "height": 269
+ },
+ "id": "rqcALNxCwKDA",
+ "outputId": "44805a78-24b9-4b41-d41a-2c8c613c87a1"
+ },
+ "source": [
+ "fig, ax = plt.subplots()\n",
+ "ax.set_xticks([])\n",
+ "ax.set_yticks([])\n",
+ "ax.imshow(true_image)"
+ ],
+ "execution_count": 20,
+ "outputs": [
+ {
+ "output_type": "execute_result",
+ "data": {
+ "text/plain": [
+ "