Skip to content

Implement IV 2SLS basic code #21

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
myeongseok-gwon merged 1 commit into from
Oct 18, 2025
Merged

Implement IV 2SLS basic code #21

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
213 changes: 212 additions & 1 deletion book/ate/iv.ipynb
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -16,6 +16,217 @@
"- Weak IV, DML 응용 등 (causal-ml book에 다양한 상황이 제시되어 있음. 실용적인 내용은 최대한 다루기)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"% pip install linearmodels"
]
},
{
"cell_type": "code",
"execution_count": 3,
"metadata": {},
"outputs": [],
"source": [
"import pandas as pd\n",
"import numpy as np\n",
"from linearmodels.iv import IV2SLS"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"push delivered(푸시 메세지 전달)와 In-App 구매력 사이의 연관관계는 인과관계가 될 수 없습니다. 소득이 confouder로 작용하기 때문입니다. (부유한 고객은 최신 스마트폰을 가져 푸시 메세지를 잘 받고, 동시에 In-App 구매력도 높기 때문입니다.)\n",
"\n",
"IV를 사용할 때, exclusion restriction가 반드시 필요합니다. 이는 정량적으로 검증할 수 없지만, 이 경우에 대해서는 push assigned(푸시 할당)는 랜덤 할당이고 다른 채널이 없기 때문에 exclusion restriction을 쉽게 주장할 수 있습니다. 다시 말해, Push Assigned는 반드시 Push Delivered를 통해서만 구매에 영향을 미칩니다."
]
},
{
"cell_type": "code",
"execution_count": 6,
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"<div>\n",
"<style scoped>\n",
" .dataframe tbody tr th:only-of-type {\n",
" vertical-align: middle;\n",
" }\n",
"\n",
" .dataframe tbody tr th {\n",
" vertical-align: top;\n",
" }\n",
"\n",
" .dataframe thead th {\n",
" text-align: right;\n",
" }\n",
"</style>\n",
"<table border=\"1\" class=\"dataframe\">\n",
" <thead>\n",
" <tr style=\"text-align: right;\">\n",
" <th></th>\n",
" <th>in_app_purchase</th>\n",
" <th>push_assigned</th>\n",
" <th>push_delivered</th>\n",
" </tr>\n",
" </thead>\n",
" <tbody>\n",
" <tr>\n",
" <th>0</th>\n",
" <td>47</td>\n",
" <td>1</td>\n",
" <td>1</td>\n",
" </tr>\n",
" <tr>\n",
" <th>1</th>\n",
" <td>43</td>\n",
" <td>1</td>\n",
" <td>0</td>\n",
" </tr>\n",
" <tr>\n",
" <th>2</th>\n",
" <td>51</td>\n",
" <td>1</td>\n",
" <td>1</td>\n",
" </tr>\n",
" <tr>\n",
" <th>3</th>\n",
" <td>49</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" </tr>\n",
" <tr>\n",
" <th>4</th>\n",
" <td>79</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" </tr>\n",
" </tbody>\n",
"</table>\n",
"</div>"
],
"text/plain": [
" in_app_purchase push_assigned push_delivered\n",
"0 47 1 1\n",
"1 43 1 0\n",
"2 51 1 1\n",
"3 49 0 0\n",
"4 79 0 0"
]
},
"execution_count": 6,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"data = pd.read_csv(\"../data/matheus_data/app_engagement_push.csv\")\n",
"data.head()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### 1st Stage (Relevance)"
]
},
{
"cell_type": "code",
"execution_count": 14,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"1st Stage: Push Assignment -> Push Delivered\n",
" Parameter Estimates \n",
"=================================================================================\n",
" Parameter Std. Err. T-stat P-value Lower CI Upper CI\n",
"---------------------------------------------------------------------------------\n",
"Intercept 2.22e-16 \n",
"push_assigned 0.7176 0.0064 112.07 0.0000 0.7050 0.7301\n",
"=================================================================================\n",
"Compliance rate: 71.76%\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"/Users/sanakang/anaconda3/lib/python3.11/site-packages/linearmodels/iv/results.py:198: RuntimeWarning: invalid value encountered in sqrt\n",
" std_errors = sqrt(diag(self.cov))\n"
]
}
],
"source": [
"print(\"1st Stage: Push Assignment -> Push Delivered\")\n",
"first_stage = IV2SLS.from_formula(\"push_delivered ~ 1 + push_assigned\", data).fit()\n",
"print(first_stage.summary.tables[1])\n",
"print(f\"Compliance rate: {first_stage.params['push_assigned']:.2%}\")"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### 2SLS Estimation: LATE"
]
},
{
"cell_type": "code",
"execution_count": 17,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
" Parameter Estimates \n",
"==================================================================================\n",
" Parameter Std. Err. T-stat P-value Lower CI Upper CI\n",
"----------------------------------------------------------------------------------\n",
"Intercept 69.292 0.3624 191.22 0.0000 68.581 70.002\n",
"push_delivered 3.2938 0.7165 4.5974 0.0000 1.8896 4.6981\n",
"==================================================================================\n"
]
}
],
"source": [
"iv_model = IV2SLS.from_formula(\"in_app_purchase ~ 1 + [push_delivered ~ push_assigned]\", data).fit()\n",
"print(iv_model.summary.tables[1])"
]
},
{
"cell_type": "code",
"execution_count": 18,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"LATE 추정치: 3.294\n",
"95% 신뢰구간: [1.890, 4.698]\n"
]
}
],
"source": [
"late_estimate = iv_model.params['push_delivered']\n",
"ci_lower = late_estimate - 1.96 * iv_model.std_errors['push_delivered'] \n",
"ci_upper = late_estimate + 1.96 * iv_model.std_errors['push_delivered']\n",
"\n",
"print(f\"LATE 추정치: {late_estimate:.3f}\")\n",
"print(f\"95% 신뢰구간: [{ci_lower:.3f}, {ci_upper:.3f}]\")"
]
},
{
"cell_type": "markdown",
"metadata": {},
Expand Down Expand Up @@ -47,7 +258,7 @@
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.12.7"
"version": "3.11.4"
}
},
"nbformat": 4,
Expand Down