From ac40d1d001de0f7e8cb943956744c33d41d941f3 Mon Sep 17 00:00:00 2001
From: = <=>
Date: Tue, 16 Jul 2024 09:03:44 -0400
Subject: [PATCH 1/8] Removed versioning

---
 .github/workflows/pr.yaml   | 21 ---------------------
 .github/workflows/push.yaml | 31 -------------------------------
 setup.py                    |  2 +-
 3 files changed, 1 insertion(+), 53 deletions(-)

diff --git a/.github/workflows/pr.yaml b/.github/workflows/pr.yaml
index 5a68682..cc39c6e 100755
--- a/.github/workflows/pr.yaml
+++ b/.github/workflows/pr.yaml
@@ -11,27 +11,6 @@ jobs:
         uses: "lgeiger/black-action@master"
         with:
           args: ". -l 79 --check"
-  check-version:
-    name: Check version
-    runs-on: ubuntu-latest
-    steps:
-      - uses: actions/checkout@v3
-        with:
-          fetch-depth: 0
-          repository: ${{ github.event.pull_request.head.repo.full_name }}
-          ref: ${{ github.event.pull_request.head.ref }}
-      - name: Set up Python
-        uses: actions/setup-python@v4
-        with:
-          python-version: 3.9
-      - name: Build changelog
-        run: pip install yaml-changelog>=0.1.7 && make changelog
-      - name: Make scripts executable
-        run: chmod -R +x .github/
-      - name: Preview changelog update
-        run: ".github/get-changelog-diff.sh"
-      - name: Check version number has been properly updated
-        run: .github/is-version-number-acceptable.sh
   Test:
     runs-on: ${{ matrix.os }}
     continue-on-error: true
diff --git a/.github/workflows/push.yaml b/.github/workflows/push.yaml
index da0d25f..2a83ac8 100755
--- a/.github/workflows/push.yaml
+++ b/.github/workflows/push.yaml
@@ -14,37 +14,6 @@ jobs:
         uses: "lgeiger/black-action@master"
         with:
           args: ". -l 79 --check"
-  versioning:
-    name: Update versioning
-    if: |
-      (github.repository == 'PolicyEngine/reweight')
-      && !(github.event.head_commit.message == 'Update reweight')
-    runs-on: ubuntu-latest
-    steps:
-      - name: Checkout repo
-        uses: actions/checkout@v3
-        with:
-          repository: ${{ github.event.pull_request.head.repo.full_name }}
-          ref: ${{ github.event.pull_request.head.ref }}
-          token: ${{ secrets.POLICYENGINE_GITHUB }}
-      - name: Setup Python
-        uses: actions/setup-python@v4
-        with:
-          python-version: 3.9
-      - name: Build changelog
-        run: pip install yaml-changelog && make changelog
-      - name: Make scripts executable
-        run: chmod -R +x .github/
-      - name: Preview changelog update
-        run: ".github/get-changelog-diff.sh"
-      - name: Update changelog
-        uses: EndBug/add-and-commit@v9
-        with:
-          add: "."
-          committer_name: Github Actions[bot]
-          author_name: Github Actions[bot]
-          message: Update reweight
-          github_token: ${{ secrets.POLICYENGINE_GITHUB }}
   Test:
     runs-on: ${{ matrix.os }}
     if: |
diff --git a/setup.py b/setup.py
index 882a501..363e940 100644
--- a/setup.py
+++ b/setup.py
@@ -7,7 +7,7 @@
 
 setup(
     name="reweight",
-    version="0.2.0",
+    version="0.3.0",
     author="PolicyEngine",
     author_email="hello@policyengine.org",
     long_description=readme,

From f6ec26c8d94ab6f087c58384110e03f67db61c76 Mon Sep 17 00:00:00 2001
From: = <=>
Date: Wed, 17 Jul 2024 10:43:17 -0400
Subject: [PATCH 2/8] Rewrote some comments in the testing notebook

---
 test.ipynb | 4 +---
 1 file changed, 1 insertion(+), 3 deletions(-)

diff --git a/test.ipynb b/test.ipynb
index 58f43c3..c219280 100644
--- a/test.ipynb
+++ b/test.ipynb
@@ -86,8 +86,6 @@
     "    # 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",
@@ -119,7 +117,7 @@
     "        # Update weights\n",
     "        optimizer.step()\n",
     "\n",
-    "        # Print loss for every 1000 epochs\n",
+    "        # Print loss for every 100 epochs\n",
     "        if epoch % 100 == 0:\n",
     "            print(f\"Epoch {epoch}, Loss: {loss.item()}\")\n",
     "\n",

From 624214d5b8718c85fa9e6a0be180ee33bcc1e2b5 Mon Sep 17 00:00:00 2001
From: = <=>
Date: Wed, 17 Jul 2024 10:43:52 -0400
Subject: [PATCH 3/8] Added (a draft of the) reweight.py function

---
 reweight/logic/reweight.py | 62 ++++++++++++++++++++++++++++++++++++++
 1 file changed, 62 insertions(+)
 create mode 100644 reweight/logic/reweight.py

diff --git a/reweight/logic/reweight.py b/reweight/logic/reweight.py
new file mode 100644
index 0000000..284cc6c
--- /dev/null
+++ b/reweight/logic/reweight.py
@@ -0,0 +1,62 @@
+import pandas as pd
+import numpy as np
+import torch
+from torch.utils.tensorboard import SummaryWriter
+
+def reweight(initial_weights, estimate_matrix, target_names, target_values, epochs = 1000, epoch_step = 100):
+    """
+    Main reweighting function, suitable for PolicyEngine UK use (PolicyEngine US use and testing TK)
+    
+    To avoid the need for equivalisation factors, use relative error:
+    |predicted - actual|/actual
+
+    Parameters:
+    household_weights (torch.Tensor): The initial weights given to survey data, which are to be
+    adjusted by this function.
+    estimate_matrix (torch.Tensor): A large matrix of estimates, obtained from e.g. a PolicyEngine
+    Microsimulation instance.
+    target_names (iterable): The names of a set of target statistics treated as ground truth.
+    target_values (torch.Tensor): The values of these target statistics.
+    epochs: The number of iterations that the optimization loop should run for.
+    epoch_step: The interval at which to print the loss during the optimization loop.
+
+    Returns:
+    final_weights: a reweighted set of household weights, obtained through an optimization process
+    over mean squared errors with respect to the target values.
+    """
+    # Initialize a TensorBoard writer
+    writer = SummaryWriter()
+
+    #Create a Torch tensor of log weights
+    log_weights = torch.log(initial_weights)
+    log_weights.requires_grad_()
+
+    # estimate_matrix (cross) exp(log_weights) = target_values
+    
+    optimizer = torch.optim.Adam([log_weights])
+
+    # Training loop
+    for epoch in range(epochs):
+
+        # Estimate the targets
+        targets_estimate = torch.exp(log_weights) @ estimate_matrix
+        # Calculate the loss
+        loss = torch.mean(((targets_estimate - target_values)/target_values) ** 2)
+
+        writer.add_scalar("Loss/train", loss, epoch)
+
+        optimizer.zero_grad()
+
+        # Perform backpropagation
+        loss.backward()
+
+        # Update weights
+        optimizer.step()
+
+        # Print loss whenever the epoch number, when one-indexed, is divisible by epoch_step
+        if (epoch+1) % epoch_step == 0:
+            print(f"Epoch {epoch+1}, Loss: {loss.item()}")
+
+    writer.flush()
+
+    return torch.exp(log_weights.detach())
\ No newline at end of file

From b4d3bc9d072861278b5286d10f6619646d6d17de Mon Sep 17 00:00:00 2001
From: = <=>
Date: Fri, 19 Jul 2024 06:08:48 -0400
Subject: [PATCH 4/8] Added more general pycache specification to gitignore

---
 .gitignore | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/.gitignore b/.gitignore
index a57fb6d..11e066e 100644
--- a/.gitignore
+++ b/.gitignore
@@ -14,6 +14,6 @@ reweight/tests/__pycache__
 #############################
 docs/_build
 
-# Pycache folder #
+# Pycache folders #
 ##################
-reweight/__pycache__
\ No newline at end of file
+**/__pycache__/
\ No newline at end of file

From 132a0041534c7ea7f31cd7c8b56849ddaa0cfb22 Mon Sep 17 00:00:00 2001
From: = <=>
Date: Mon, 22 Jul 2024 09:55:00 -0400
Subject: [PATCH 5/8] Added a new test for the reweight logic

---
 reweight/tests/test_uk_prototype.py | 22 ++++++++++++++++++++++
 1 file changed, 22 insertions(+)

diff --git a/reweight/tests/test_uk_prototype.py b/reweight/tests/test_uk_prototype.py
index 35ffbce..1265073 100644
--- a/reweight/tests/test_uk_prototype.py
+++ b/reweight/tests/test_uk_prototype.py
@@ -3,3 +3,25 @@ def test_uk_microsimulation():
 
     # Create a Microsimulation instance
     sim = Microsimulation()
+
+def test_uk_prototype():
+    from policyengine_uk import Microsimulation
+    from reweight.logic import reweight
+    import torch
+    sim = Microsimulation()
+    
+    from policyengine_uk.data import RawFRS_2021_22
+    RawFRS_2021_22().download()
+
+    from policyengine_uk.data.datasets.frs.calibration.calibrate import generate_model_variables
+    (
+        household_weights,
+        weight_adjustment,
+        values_df,
+        targets,
+        targets_array,
+        equivalisation_factors_array
+    ) = generate_model_variables("frs_2021", 2025)
+
+    sim_matrix = torch.tensor(values_df.to_numpy(), dtype=torch.float32)
+    reweight.reweight(household_weights, sim_matrix, targets, targets_array)
\ No newline at end of file

From fb081962279a1bb9e45586d547701ea2fbf9d0dd Mon Sep 17 00:00:00 2001
From: = <=>
Date: Mon, 22 Jul 2024 10:10:27 -0400
Subject: [PATCH 6/8] Reformatted

---
 reweight/logic/reweight.py          | 24 +++++++++++++++++-------
 reweight/tests/test_uk_prototype.py | 14 ++++++++++----
 2 files changed, 27 insertions(+), 11 deletions(-)

diff --git a/reweight/logic/reweight.py b/reweight/logic/reweight.py
index 284cc6c..859e962 100644
--- a/reweight/logic/reweight.py
+++ b/reweight/logic/reweight.py
@@ -3,10 +3,18 @@
 import torch
 from torch.utils.tensorboard import SummaryWriter
 
-def reweight(initial_weights, estimate_matrix, target_names, target_values, epochs = 1000, epoch_step = 100):
+
+def reweight(
+    initial_weights,
+    estimate_matrix,
+    target_names,
+    target_values,
+    epochs=1000,
+    epoch_step=100,
+):
     """
     Main reweighting function, suitable for PolicyEngine UK use (PolicyEngine US use and testing TK)
-    
+
     To avoid the need for equivalisation factors, use relative error:
     |predicted - actual|/actual
 
@@ -27,12 +35,12 @@ def reweight(initial_weights, estimate_matrix, target_names, target_values, epoc
     # Initialize a TensorBoard writer
     writer = SummaryWriter()
 
-    #Create a Torch tensor of log weights
+    # Create a Torch tensor of log weights
     log_weights = torch.log(initial_weights)
     log_weights.requires_grad_()
 
     # estimate_matrix (cross) exp(log_weights) = target_values
-    
+
     optimizer = torch.optim.Adam([log_weights])
 
     # Training loop
@@ -41,7 +49,9 @@ def reweight(initial_weights, estimate_matrix, target_names, target_values, epoc
         # Estimate the targets
         targets_estimate = torch.exp(log_weights) @ estimate_matrix
         # Calculate the loss
-        loss = torch.mean(((targets_estimate - target_values)/target_values) ** 2)
+        loss = torch.mean(
+            ((targets_estimate - target_values) / target_values) ** 2
+        )
 
         writer.add_scalar("Loss/train", loss, epoch)
 
@@ -54,9 +64,9 @@ def reweight(initial_weights, estimate_matrix, target_names, target_values, epoc
         optimizer.step()
 
         # Print loss whenever the epoch number, when one-indexed, is divisible by epoch_step
-        if (epoch+1) % epoch_step == 0:
+        if (epoch + 1) % epoch_step == 0:
             print(f"Epoch {epoch+1}, Loss: {loss.item()}")
 
     writer.flush()
 
-    return torch.exp(log_weights.detach())
\ No newline at end of file
+    return torch.exp(log_weights.detach())
diff --git a/reweight/tests/test_uk_prototype.py b/reweight/tests/test_uk_prototype.py
index 1265073..b413f2f 100644
--- a/reweight/tests/test_uk_prototype.py
+++ b/reweight/tests/test_uk_prototype.py
@@ -4,24 +4,30 @@ def test_uk_microsimulation():
     # Create a Microsimulation instance
     sim = Microsimulation()
 
+
 def test_uk_prototype():
     from policyengine_uk import Microsimulation
     from reweight.logic import reweight
     import torch
+
     sim = Microsimulation()
-    
+
     from policyengine_uk.data import RawFRS_2021_22
+
     RawFRS_2021_22().download()
 
-    from policyengine_uk.data.datasets.frs.calibration.calibrate import generate_model_variables
+    from policyengine_uk.data.datasets.frs.calibration.calibrate import (
+        generate_model_variables,
+    )
+
     (
         household_weights,
         weight_adjustment,
         values_df,
         targets,
         targets_array,
-        equivalisation_factors_array
+        equivalisation_factors_array,
     ) = generate_model_variables("frs_2021", 2025)
 
     sim_matrix = torch.tensor(values_df.to_numpy(), dtype=torch.float32)
-    reweight.reweight(household_weights, sim_matrix, targets, targets_array)
\ No newline at end of file
+    reweight.reweight(household_weights, sim_matrix, targets, targets_array)

From 349d70772594c0406b5fef2ee232238ed253d9ae Mon Sep 17 00:00:00 2001
From: = <=>
Date: Mon, 22 Jul 2024 10:58:25 -0400
Subject: [PATCH 7/8] Added init scripts to reweight directories

---
 changelog_entry.yaml                | 1 +
 reweight/data/__init__.py           | 1 +
 reweight/logic/__init__.py          | 1 +
 reweight/tests/__init__.py          | 1 +
 reweight/tests/test_uk_prototype.py | 2 +-
 5 files changed, 5 insertions(+), 1 deletion(-)
 create mode 100644 reweight/data/__init__.py
 create mode 100644 reweight/logic/__init__.py
 create mode 100644 reweight/tests/__init__.py

diff --git a/changelog_entry.yaml b/changelog_entry.yaml
index e69de29..8b13789 100644
--- a/changelog_entry.yaml
+++ b/changelog_entry.yaml
@@ -0,0 +1 @@
+
diff --git a/reweight/data/__init__.py b/reweight/data/__init__.py
new file mode 100644
index 0000000..badcc11
--- /dev/null
+++ b/reweight/data/__init__.py
@@ -0,0 +1 @@
+from .reweight import reweight
\ No newline at end of file
diff --git a/reweight/logic/__init__.py b/reweight/logic/__init__.py
new file mode 100644
index 0000000..badcc11
--- /dev/null
+++ b/reweight/logic/__init__.py
@@ -0,0 +1 @@
+from .reweight import reweight
\ No newline at end of file
diff --git a/reweight/tests/__init__.py b/reweight/tests/__init__.py
new file mode 100644
index 0000000..badcc11
--- /dev/null
+++ b/reweight/tests/__init__.py
@@ -0,0 +1 @@
+from .reweight import reweight
\ No newline at end of file
diff --git a/reweight/tests/test_uk_prototype.py b/reweight/tests/test_uk_prototype.py
index b413f2f..bb08678 100644
--- a/reweight/tests/test_uk_prototype.py
+++ b/reweight/tests/test_uk_prototype.py
@@ -7,7 +7,7 @@ def test_uk_microsimulation():
 
 def test_uk_prototype():
     from policyengine_uk import Microsimulation
-    from reweight.logic import reweight
+    from reweight import reweight
     import torch
 
     sim = Microsimulation()

From 7543011bdcce96987fd539a0ab3cccd83c3c9f50 Mon Sep 17 00:00:00 2001
From: = <=>
Date: Mon, 22 Jul 2024 16:11:30 -0400
Subject: [PATCH 8/8] Added __init__ scripts for logic directory

---
 reweight/__init__.py                |  3 +-
 reweight/data/__init__.py           |  1 -
 reweight/logic/__init__.py          |  2 +-
 reweight/tests/__init__.py          |  1 -
 reweight/tests/test_uk_prototype.py |  2 +-
 test.ipynb                          | 48 +++++++++++++++++++++++++++++
 6 files changed, 52 insertions(+), 5 deletions(-)
 delete mode 100644 reweight/data/__init__.py
 delete mode 100644 reweight/tests/__init__.py

diff --git a/reweight/__init__.py b/reweight/__init__.py
index 3dc1f76..28fe9cc 100644
--- a/reweight/__init__.py
+++ b/reweight/__init__.py
@@ -1 +1,2 @@
-__version__ = "0.1.0"
+__version__ = "0.3.0"
+from .logic.reweight import reweight
diff --git a/reweight/data/__init__.py b/reweight/data/__init__.py
deleted file mode 100644
index badcc11..0000000
--- a/reweight/data/__init__.py
+++ /dev/null
@@ -1 +0,0 @@
-from .reweight import reweight
\ No newline at end of file
diff --git a/reweight/logic/__init__.py b/reweight/logic/__init__.py
index badcc11..a9de83b 100644
--- a/reweight/logic/__init__.py
+++ b/reweight/logic/__init__.py
@@ -1 +1 @@
-from .reweight import reweight
\ No newline at end of file
+from .reweight import reweight
diff --git a/reweight/tests/__init__.py b/reweight/tests/__init__.py
deleted file mode 100644
index badcc11..0000000
--- a/reweight/tests/__init__.py
+++ /dev/null
@@ -1 +0,0 @@
-from .reweight import reweight
\ No newline at end of file
diff --git a/reweight/tests/test_uk_prototype.py b/reweight/tests/test_uk_prototype.py
index bb08678..dc69b2b 100644
--- a/reweight/tests/test_uk_prototype.py
+++ b/reweight/tests/test_uk_prototype.py
@@ -30,4 +30,4 @@ def test_uk_prototype():
     ) = generate_model_variables("frs_2021", 2025)
 
     sim_matrix = torch.tensor(values_df.to_numpy(), dtype=torch.float32)
-    reweight.reweight(household_weights, sim_matrix, targets, targets_array)
+    reweight(household_weights, sim_matrix, targets, targets_array)
diff --git a/test.ipynb b/test.ipynb
index c219280..e346cd3 100644
--- a/test.ipynb
+++ b/test.ipynb
@@ -76,6 +76,54 @@
     "# square error, and then average to get MSE."
    ]
   },
+  {
+   "cell_type": "code",
+   "execution_count": 9,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "from reweight.logic import reweight"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 11,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 100, Loss: 48.30685043334961\n",
+      "Epoch 200, Loss: 40.58155059814453\n",
+      "Epoch 300, Loss: 34.585235595703125\n",
+      "Epoch 400, Loss: 29.832853317260742\n",
+      "Epoch 500, Loss: 25.99891471862793\n",
+      "Epoch 600, Loss: 22.858182907104492\n",
+      "Epoch 700, Loss: 20.250896453857422\n",
+      "Epoch 800, Loss: 18.061073303222656\n",
+      "Epoch 900, Loss: 16.202829360961914\n",
+      "Epoch 1000, Loss: 14.611446380615234\n"
+     ]
+    },
+    {
+     "data": {
+      "text/plain": [
+       "tensor([1120.1953,   89.4442, 3851.2649,  ...,  730.9640,  832.0632,\n",
+       "        4155.2686])"
+      ]
+     },
+     "execution_count": 11,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "sim_matrix = torch.tensor(values_df.to_numpy(), dtype=torch.float32)\n",
+    "\n",
+    "reweight.reweight(household_weights, sim_matrix, targets, targets_array)"
+   ]
+  },
   {
    "cell_type": "code",
    "execution_count": 29,