def get_common_dataframe(variables, year=2006):
"""
Compare variables in erf an openfisca
"""
simulation = SurveySimulation()
simulation.set_config(year=year)
simulation.set_param()
simulation.set_survey()
simulation.compute()
erf = ErfsDataTable(year=2006)
if "ident" not in variables:
erf_variables = variables + ["ident"]
else:
erf_variables = variables
if "wprm" not in erf_variables:
erf_variables = erf_variables + ["wprm"]
else:
erf_variables = erf_variables
erf_dataframe = erf.get_values(erf_variables, table="menage")
erf_dataframe.rename(columns={'ident': 'idmen'}, inplace=True)
for col in erf_dataframe.columns:
if col is not "idmen":
erf_dataframe.rename(columns={col: col + "_erf"}, inplace=True)
of_dataframe, of_dataframe_default = simulation.aggregated_by_entity(
"men", variables, all_output_vars=False, force_sum=True)
del of_dataframe_default
merged_df = of_dataframe.merge(erf_dataframe, on="idmen")
del of_dataframe, erf_dataframe
return merged_df
def get_menage(idmen):
erf = ErfsDataTable(year=2006)
erf_variables = [
"ident", "age", "noi", "persfip", "persfipd", "quelfic", "declar1",
"declar2"
]
dataframe = erf.get_values(erf_variables, table="indivi")
dataframe.rename(columns={'ident': 'idmen'}, inplace=True)
selection = dataframe.idmen == idmen
return dataframe[selection]
# This file is part of OpenFisca.
# OpenFisca is a socio-fiscal microsimulation software
# Copyright © #2013 Clément Schaff, Mahdi Ben Jelloul
# Licensed under the terms of the GVPLv3 or later license
# (see openfisca/__init__.py for details)
from numpy import arange
from openfisca_core.statshelpers import mark_weighted_percentiles
from openfisca_france.data.erf.datatable import ErfsDataTable
from openfisca_france.data.sources.config import destination_dir
from pandas import DataFrame, ExcelWriter
year = 2009
erf = ErfsDataTable(year=year)
#erf.set_config()
vars = [ "zrstm", "zchom", "pfamm", "wprm", "pauvre50m", "pauvre60m", "nivviem", "champm"]
df = erf.get_values(variables=vars, table="erf_menage" )
labels = arange(1,11)
method = 2
nivvie = df["nivviem"].astype("float64").values
wprm = df["wprm"].astype("float64").values
decil, values = mark_weighted_percentiles(nivvie, labels, wprm, method, return_quantiles = True)
df2 = DataFrame({"decile" : decil})