Merge pull request #11 from PolicyEngine/uk-addition

Adding PolicyEngine UK functionality, including microdata testing
PolicyEngine · Jul 16, 2024 · e43d080 · e43d080
2 parents 1d39b67 + b513ff3
commit e43d080
Show file tree

Hide file tree

Showing 9 changed files with 225 additions and 5 deletions.
diff --git a/.github/workflows/pr.yaml b/.github/workflows/pr.yaml
@@ -51,6 +51,7 @@ jobs:
         run: make test
         env:
           POVERTYTRACKER_RAW_URL: ${{ secrets.POVERTYTRACKER_RAW_URL }}
+          POLICYENGINE_GITHUB_MICRODATA_AUTH_TOKEN: ${{ secrets.POLICYENGINE_GITHUB_MICRODATA_AUTH_TOKEN}}
       - uses: codecov/codecov-action@v3
       - name: Build package
         run: make

diff --git a/.github/workflows/push.yaml b/.github/workflows/push.yaml
@@ -66,6 +66,7 @@ jobs:
         run: make test
         env:
           POVERTYTRACKER_RAW_URL: ${{ secrets.POVERTYTRACKER_RAW_URL }}
+          POLICYENGINE_GITHUB_MICRODATA_AUTH_TOKEN: ${{ secrets.POLICYENGINE_GITHUB_MICRODATA_AUTH_TOKEN}}
       - uses: codecov/codecov-action@v3
       - name: Test documentation builds
         if: matrix.os == 'ubuntu-latest'

diff --git a/CHANGELOG.md b/CHANGELOG.md
@@ -5,6 +5,12 @@ All notable changes to this project will be documented in this file.
 The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/),
 and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
 
+## [0.3.0] - 2024-07-16 07:44:29
+
+### Added
+
+- PolicyEngine UK testing
+
 ## [0.2.0] - 2024-07-02 13:04:56
 
 ### Added
@@ -17,7 +23,5 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 
 - Basic repo
 
-
-
+[1.0.0]: https://github.com/PolicyEngine/reweight/compare/0.2.0...1.0.0
 [0.2.0]: https://github.com/PolicyEngine/reweight/compare/0.1.0...0.2.0
-
diff --git a/changelog.yaml b/changelog.yaml
@@ -1,10 +1,15 @@
 - changes:
     added:
-    - Basic repo
+      - Basic repo
   date: 2024-06-26 00:00:00
   version: 0.1.0
 - bump: minor
   changes:
     added:
-    - PyTorch testing with Microsimulation datasets
+      - PyTorch testing with Microsimulation datasets
   date: 2024-07-02 13:04:56
+- bump: minor
+  changes:
+    added:
+      - PolicyEngine UK testing
+  date: 2024-07-16 07:44:29
diff --git a/changelog_entry.yaml b/changelog_entry.yaml
@@ -0,0 +1,4 @@
+- bump: minor
+  changes:
+    fixed:
+      - PolicyEngine UK secrets issues
diff --git a/reweight/tests/test_installation.py b/reweight/tests/test_installation.py
@@ -13,3 +13,11 @@ def test_install():
         import reweight
     except:
         raise AssertionError("Failed to build reweight")
+
+
+def test_secret_usage():
+    import os
+
+    token = os.environ["POLICYENGINE_GITHUB_MICRODATA_AUTH_TOKEN"]
+    token_not_none = token != None
+    assert token_not_none, "Authentication token is None"
diff --git a/reweight/tests/test_uk_prototype.py b/reweight/tests/test_uk_prototype.py
@@ -0,0 +1,5 @@
+def test_uk_microsimulation():
+    from policyengine_uk import Microsimulation
+
+    # Create a Microsimulation instance
+    sim = Microsimulation()
diff --git a/setup.py b/setup.py
@@ -34,6 +34,7 @@
         "pytest",
         "policyengine-core~=2.21.8",
         "policyengine-us~=0.794.1",
+        "policyengine-uk",
     ],
     extras_require={
         "dev": [

diff --git a/test.ipynb b/test.ipynb
@@ -0,0 +1,191 @@
+{
+ "cells": [
+  {
+   "cell_type": "code",
+   "execution_count": 1,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "from policyengine_uk import Microsimulation"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 2,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "import os\n",
+    "token = os.environ[\"POLICYENGINE_GITHUB_MICRODATA_AUTH_TOKEN\"]"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 3,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "sim = Microsimulation()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 4,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "from policyengine_uk.data import RawFRS_2021_22\n",
+    "RawFRS_2021_22().download()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 5,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "from policyengine_uk.data.datasets.frs.calibration.calibrate import generate_model_variables\n",
+    "\n",
+    "(\n",
+    "    household_weights,\n",
+    "    weight_adjustment,\n",
+    "    values_df,\n",
+    "    targets,\n",
+    "    targets_array,\n",
+    "    equivalisation_factors_array\n",
+    ") = generate_model_variables(\"frs_2021\", 2025)\n",
+    "\n",
+    "#This returns a set of household weights, a random tensor of the same size as the weights tensor,\n",
+    "#a Pandas dataframe to transform weights into statistical predictions, a dictionary of target values,\n",
+    "#an array of target values, and some equivalisation factors I don't understand."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 6,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "import pandas as pd\n",
+    "import numpy as np\n",
+    "import torch\n",
+    "from torch.utils.tensorboard import SummaryWriter\n",
+    "\n",
+    "# Then we're working with: this new array * the weights = our estimate.\n",
+    "# Then our error in a prediction is based on |predicted - actual|/equivalisation factor. Square that to get\n",
+    "# square error, and then average to get MSE."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 29,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "def calibrate(household_weights, weight_adjustment, values_df, targets, targets_array, equivalisation_factors_array):\n",
+    "    # Initialize a TensorBoard writer\n",
+    "    writer = SummaryWriter()\n",
+    "\n",
+    "    #TODO: Write stuff here\n",
+    "\n",
+    "    #Create a Torch tensor of log weights\n",
+    "    log_weights = torch.log(household_weights)\n",
+    "    log_weights.requires_grad_()\n",
+    "\n",
+    "    sim_matrix = values_df.to_numpy()\n",
+    "\n",
+    "    # sim_matrix (cross) exp(log_weights) = targets_array\n",
+    "    sim_matrix = torch.tensor(sim_matrix, dtype=torch.float32)\n",
+    "    #targets_array will be our target values.\n",
+    "\n",
+    "    optimizer = torch.optim.Adam([log_weights])\n",
+    "\n",
+    "    # Training loop\n",
+    "    num_epochs = 1000\n",
+    "    for epoch in range(num_epochs):\n",
+    "\n",
+    "        # Estimate the targets\n",
+    "        targets_estimate = torch.exp(log_weights) @ sim_matrix\n",
+    "        # Calculate the loss\n",
+    "        loss = torch.mean(((targets_estimate - targets_array)/equivalisation_factors_array) ** 2)\n",
+    "\n",
+    "        writer.add_scalar(\"Loss/train\", loss, epoch)\n",
+    "\n",
+    "        optimizer.zero_grad()\n",
+    "\n",
+    "        # Perform backpropagation\n",
+    "        loss.backward()\n",
+    "\n",
+    "        # Update weights\n",
+    "        optimizer.step()\n",
+    "\n",
+    "        # Print loss for every 1000 epochs\n",
+    "        if epoch % 100 == 0:\n",
+    "            print(f\"Epoch {epoch}, Loss: {loss.item()}\")\n",
+    "\n",
+    "    writer.flush()\n",
+    "\n",
+    "    final_weights = np.exp(log_weights.detach().numpy())\n",
+    "    final_estimates = (\n",
+    "        final_weights @ sim_matrix.numpy()\n",
+    "    )\n",
+    "    true_values = targets\n",
+    "    #print(\"Final weights:\", final_weights)\n",
+    "    #print(\"Final estimates:\", final_estimates)\n",
+    "    #print(\"True values:\", true_values)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 30,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 0, Loss: 0.2274675965309143\n",
+      "Epoch 100, Loss: 0.18678854405879974\n",
+      "Epoch 200, Loss: 0.15837892889976501\n",
+      "Epoch 300, Loss: 0.13632304966449738\n",
+      "Epoch 400, Loss: 0.11881797015666962\n",
+      "Epoch 500, Loss: 0.1046074628829956\n",
+      "Epoch 600, Loss: 0.09283030778169632\n",
+      "Epoch 700, Loss: 0.08289289474487305\n",
+      "Epoch 800, Loss: 0.0743781179189682\n",
+      "Epoch 900, Loss: 0.06698659062385559\n"
+     ]
+    }
+   ],
+   "source": [
+    "calibrate(household_weights,\n",
+    "    weight_adjustment,\n",
+    "    values_df,\n",
+    "    targets,\n",
+    "    targets_array,\n",
+    "    equivalisation_factors_array)"
+   ]
+  }
+ ],
+ "metadata": {
+  "kernelspec": {
+   "display_name": "policyengine",
+   "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": "3.9.7"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 2
+}