def attach_nid(df):
"""
add merged (private + medicare supplemental insurance) NIDs
onto the marketscan data
"""
nid_dict = {2000: 244369,
2010: 244370,
2011: 336850,
2012: 244371,
2013: 336849,
2014: 336848,
2015: 336847,
2016: 408680}
df = hosp_prep.fill_nid(df, nid_dict)
assert (df['nid'] == 0).sum() == 0, "There are missing NIDs"
return df
2001: 87001,
2002: 87002,
2003: 87003,
2004: 87004,
2005: 87005,
2006: 87006,
2007: 87007,
2008: 87008,
2009: 87009,
2010: 87010,
2011: 87011,
2012: 114876,
2013: 160484,
2014: 237756
}
df = hosp_prep.fill_nid(df, nid_dictionary)
fix_sex_dict = {'Mujeres': 2, 'Mujer': 2, 'Hombres': 1, 'Hombre': 1}
df['sex_id'].replace(fix_sex_dict, inplace=True)
df['sex_id'] = pd.to_numeric(df['sex_id'], downcast='integer', errors='raise')
df.loc[(df['sex_id'] != 1) & (df['sex_id'] != 2), 'sex_id'] = 3
assert set(df.sex_id.unique()).issubset({1, 2, 3})
age_unit_translation_dict = {
"A\xf1os (1 a 98 a\xf1os de edad)": "Years",
"Meses (1 a 11 meses de edad)": "Months",
'D\xedas (1 a 29 d\xedas de edad)': 'Days',
'Ignorado': 'Unknown',
'Anios (1 a 115 anios de edad)': "Years",
nid_dict = {
2004: 86953,
2005: 86954,
2006: 86955,
2007: 86956,
2008: 86957,
2009: 86958,
2010: 94170,
2011: 94171,
2012: 121282,
2013: 150449,
2014: 220205,
2015: 281773
}
df = hosp_prep.fill_nid(df, nid_dict)
df.los.value_counts(dropna=False).head()
df.los.isnull().sum()
df = df[df['los'] > 0]
final_admits = len(df)
df['facility_id'] = 'hospital'
df = hosp_prep.age_binning(df, drop_age=True)
int_cols = [
'location_id', 'year_start', 'year_end', 'age_group_unit', 'age_start',
'age_end', 'sex_id', 'nid', 'representative_id', 'metric_id'
]