import pandas as pd
import numpy as np
from ogcore.utils import save_return_table
from ogcore.constants import VAR_LABELS, PARAM_LABELS, DEFAULT_START_YEAR
[docs]
def tax_rate_table(
base_TxFuncEst,
base_params,
reform_TxFuncEst=None,
reform_params=None,
rate_type="ETR",
start_year=DEFAULT_START_YEAR,
num_years=10,
table_format="tex",
path=None,
):
"""
Table of average tax rates over several years.
Args:
base_TxFuncEst(dictionary): Baseline tax function parameter
estimates
base_params (OG-Core Specifications class): baseline parameters
object
reform_TxFuncEst (dictionary): Reform tax function parameter
estimates
reform_params (OG-Core Specifications class): reform parameters
object
rate_type (string): Tax rate to include in table
start_year (integer): year to start table
num_years (integer): number of years to include in table
table_format (string): format to save/return table as
path (string): path to save table to
Returns:
table_str (string or DataFrame): table of tax rates
"""
assert isinstance(start_year, (int, np.integer))
assert isinstance(num_years, (int, np.integer))
# Make sure both runs cover same time period
if reform_TxFuncEst is not None:
assert base_params.start_year == reform_params.start_year
start_index = start_year - base_params.start_year
years = list(np.arange(start_year, start_year + num_years, 1))
if reform_TxFuncEst is None:
if rate_type == "ETR":
rates = base_TxFuncEst["tfunc_avg_etr"] * 100
elif rate_type == "MTRx":
rates = base_TxFuncEst["tfunc_avg_mtrx"] * 100
elif rate_type == "MTRy":
rates = base_TxFuncEst["tfunc_avg_mtry"] * 100
elif rate_type == "all":
etr_rates = base_TxFuncEst["tfunc_avg_etr"] * 100
mtrx_rates = base_TxFuncEst["tfunc_avg_mtrx"] * 100
mtry_rates = base_TxFuncEst["tfunc_avg_mtry"] * 100
else:
raise ValueError(
"Value {!r} is not a valid rate_type".format(rate_type)
)
if rate_type == "all":
# In case num_years is greater than number of years
# tax function estimates are for
len_rates = len(etr_rates[start_index : start_index + num_years])
table = {
"Year": years[:len_rates],
VAR_LABELS["ETR"]: etr_rates[
start_index : start_index + num_years
],
VAR_LABELS["MTRx"]: mtrx_rates[
start_index : start_index + num_years
],
VAR_LABELS["MTRy"]: mtry_rates[
start_index : start_index + num_years
],
}
else:
len_rates = len(rates[start_index : start_index + num_years])
table = {
"Year": years[:len_rates],
VAR_LABELS[rate_type]: rates[
start_index : start_index + num_years
],
}
else:
if rate_type == "ETR":
base_rates = base_TxFuncEst["tfunc_avg_etr"] * 100
reform_rates = reform_TxFuncEst["tfunc_avg_etr"] * 100
elif rate_type == "MTRx":
base_rates = base_TxFuncEst["tfunc_avg_mtrx"] * 100
reform_rates = reform_TxFuncEst["tfunc_avg_mtrx"] * 100
elif rate_type == "MTRy":
base_rates = base_TxFuncEst["tfunc_avg_mtrx"] * 100
reform_rates = reform_TxFuncEst["tfunc_avg_mtrx"] * 100
elif rate_type == "all":
base_etr_rates = base_TxFuncEst["tfunc_avg_etr"] * 100
base_mtrx_rates = base_TxFuncEst["tfunc_avg_mtrx"] * 100
base_mtry_rates = base_TxFuncEst["tfunc_avg_mtry"] * 100
reform_etr_rates = reform_TxFuncEst["tfunc_avg_etr"] * 100
reform_mtrx_rates = reform_TxFuncEst["tfunc_avg_mtrx"] * 100
reform_mtry_rates = reform_TxFuncEst["tfunc_avg_mtry"] * 100
else:
raise ValueError(
"Value {!r} is not a valid rate_type".format(rate_type)
)
if rate_type == "all":
len_rates = len(base_etr_rates[start_index : start_index + num_years])
table = {
"Year": years[:len_rates],
"Baseline "
+ VAR_LABELS["ETR"]: base_etr_rates[
start_index : start_index + num_years
],
"Reform "
+ VAR_LABELS["ETR"]: reform_etr_rates[
start_index : start_index + num_years
],
"Differences in "
+ VAR_LABELS["ETR"]: reform_etr_rates[
start_index : start_index + num_years
]
- base_etr_rates[start_index : start_index + num_years],
"Baseline "
+ VAR_LABELS["MTRx"]: base_mtrx_rates[
start_index : start_index + num_years
],
"Reform "
+ VAR_LABELS["MTRx"]: reform_mtrx_rates[
start_index : start_index + num_years
],
"Differences in "
+ VAR_LABELS["MTRx"]: reform_mtrx_rates[
start_index : start_index + num_years
]
- base_mtrx_rates[start_index : start_index + num_years],
"Baseline "
+ VAR_LABELS["MTRy"]: base_mtry_rates[
start_index : start_index + num_years
],
"Reform "
+ VAR_LABELS["MTRy"]: reform_mtry_rates[
start_index : start_index + num_years
],
"Differences in "
+ VAR_LABELS["MTRy"]: reform_mtry_rates[
start_index : start_index + num_years
]
- base_mtry_rates[start_index : start_index + num_years],
}
else:
len_rates = len(base_rates[start_index : start_index + num_years])
table = {
"Year": years[:len_rates],
"Baseline "
+ VAR_LABELS[rate_type]: base_rates[
start_index : start_index + num_years
],
"Reform "
+ VAR_LABELS[rate_type]: reform_rates[
start_index : start_index + num_years
],
"Difference": reform_rates[start_index : start_index + num_years]
- base_rates[start_index : start_index + num_years],
}
table_df = (pd.DataFrame.from_dict(table, orient="columns")).transpose()
table_df.columns = table_df.iloc[0].astype("int").astype("str")
table_df.reindex(table_df.index.drop("Year"))
table_df.drop("Year", inplace=True)
table_df.reset_index(inplace=True)
table_df.rename(columns={"index": "Variable"}, inplace=True)
table_str = save_return_table(table_df, table_format, path, precision=2)
return table_str
[docs]
def param_table(p, table_format="tex", path=None):
"""
This function creates a table of model parameters for publication.
Args:
p (OG-Core Specifications class): baseline parameters
object)
table_format (string): format to save/return table as
path (string): path to save table to
Returns:
table (string or DataFrame): table of tax rates
"""
table = {"Symbol": [], "Description": [], "Value": []}
for k, v in PARAM_LABELS.items():
table["Symbol"].append(v[1])
table["Description"].append(v[0])
value = getattr(p, k)
if hasattr(value, "__len__") & ~isinstance(value, str):
if value.ndim > 1:
report = (
"Too large to report here, see default parameters JSON"
)
else:
report = (
"["
+ "{0:1.3f}".format(value[0])
+ "..."
+ "{0:1.3f}".format(value[-1])
+ "]"
)
else:
if isinstance(value, int) or isinstance(value, np.int64):
report = str(value)
elif isinstance(value, str):
report = value
else:
if value < 0.0001:
report = "{:.2E}".format(value)
else:
report = "{0:1.3f}".format(value)
table["Value"].append(report)
table_df = pd.DataFrame.from_dict(table)
table_str = save_return_table(table_df, table_format, path, precision=3)
return table_str
