Assignment/ivs 219385

assignments/ivs/code/grader.py

@@ -1,8 +1,9 @@
 # Imports
-import importlib.util
+import os
 import sys
 import traceback
-import os
+import importlib.util
+import numpy as np
 
 # Find student's script
 def find_student_script():

assignments/ivs/code/ivs-219385.py

@@ -0,0 +1,52 @@
+#importar paquetes
+import pandas as pd
+import statsmodels.api as sm
+from linearmodels.iv import IV2SLS
+from itertools import product
+
+# Cargar y preparar datos
+df = pd.read_csv('../data/raw.csv')
+df = df[df['yob'] >= 1940]
+df['cohort'] = '40-49'
+
+# Crear dummies para año y trimestre
+yob_dummies = pd.get_dummies(df['yob'], prefix='yob').astype(int)
+qob_dummies = pd.get_dummies(df['qob'], prefix='qob').astype(int)
+df = pd.concat([df, yob_dummies, qob_dummies], axis=1)
+
+# Crear interacciones yob × qob (excepto _qob_4 se usará como referencia después)
+yob_cols = [col for col in df.columns if col.startswith('yob_') and '_qob_' not in col]
+qob_cols = [col for col in df.columns if col.startswith('qob_')]
+for yob, qob in product(yob_cols, qob_cols):
+    df[f'{yob}_{qob}'] = (df[yob] * df[qob]).astype(int)
+
+# Agregar constante
+df['const'] = 1
+
+# Modelo OLS (res0) 
+controls = [
+    'educ', 'race', 'married', 'smsa', 'neweng', 'midatl', 'enocent',
+    'wnocent', 'soatl', 'esocent', 'wsocent', 'mt'
+]
+yob_dummies = [col for col in df.columns if col.startswith('yob_') and '1949' not in col and 'qob_' not in col]
+X = df[['const'] + controls + yob_dummies]
+y = df['lwklywge']
+res0 = sm.OLS(y, X).fit(cov_type='HC3')
+
+# Modelo IV (res1)
+controls_iv = [
+    'race', 'married', 'smsa', 'neweng', 'midatl', 'enocent',
+    'wnocent', 'soatl', 'esocent', 'wsocent', 'mt'
+]
+yob_dummies_iv = [col for col in df.columns if col.startswith('yob_') and '1949' not in col and '_qob_' not in col]
+instr_cols = [col for col in df.columns if '_qob_' in col and not col.endswith('_qob_4')]
+
+exog = df[['const'] + controls_iv + yob_dummies_iv]
+endog = df['educ']
+instruments = df[instr_cols]
+res1 = IV2SLS(dependent=y, exog=exog, endog=endog, instruments=instruments).fit(cov_type='robust')
+
+# sesgo
+bias = True
+bias_sign = '+'
+

assignments/rct/code/rct-219385.py

@@ -0,0 +1,48 @@
+#imports
+import os
+import pandas as pd
+import statsmodels.api as sm
+
+# Load data
+PATH = os.path.join('..', 'data', 'raw.csv')
+df = pd.read_csv(PATH)
+
+## Renombrar columnas
+df.columns = ['id', 'dark', 'views', 'time', 'purchase', 'mobile', 'location']
+
+# Mapear columnas a valores numéricos
+df.replace(
+    to_replace={
+        'dark': {'A': '0', 'B': '1'},
+        'mobile': {'Mobile': '1', 'Desktop': '0'},
+        'purchase': {'No': '0', 'Yes': '1'},
+        'location': {'Northern Ireland': 'Ireland'}
+    },
+    inplace=True
+)
+
+# Convertir strings a enteros
+df[['dark', 'mobile', 'purchase']] = df[['dark', 'mobile', 'purchase']].astype(int)
+
+# Convertir la columna 'location' a minúsculas
+df['location'] = df['location'].str.lower()
+
+df = pd.get_dummies(
+    data=df,
+    prefix='',
+    prefix_sep='',
+    columns=['location'],
+    dtype=int
+)
+
+# Asignar una constante
+df['const'] = 1
+
+# Declarar modelo sin término de interacción
+spec = sm.OLS(
+    endog=df['purchase'],  # Variable dependiente
+    exog=df[['const', 'ireland', 'scotland', 'wales', 'dark']],  # Variables explicativas sin interacción
+)
+
+# Ajustar el modelo
+model = spec.fit()