def test_simple_eval():
specs = Specifications()
specs.profit_rate = 1.0
assert specs.simple_eval('profit_rate / 2') == 0.5
assert specs.simple_eval('profit_rate * 2') == 2.0
assert specs.simple_eval('profit_rate - 2') == -1.0
assert specs.simple_eval('profit_rate + 2') == 3.0
def test_simple_eval():
specs = Specifications()
specs.profit_rate = 1.0
assert specs.simple_eval('profit_rate / 2') == 0.5
assert specs.simple_eval('profit_rate * 2') == 2.0
assert specs.simple_eval('profit_rate - 2') == -1.0
assert specs.simple_eval('profit_rate + 2') == 3.0
def test_implement_reform():
specs = Specifications()
new_specs = {'profit_rate': 0.4, 'm': 0.5}
specs.update_specifications(new_specs)
assert specs.profit_rate == 0.4
assert specs.m == 0.5
assert len(specs.errors) == 0
def validate_inputs(meta_param_dict, adjustment, errors_warnings):
'''
Validates user inputs for parameters
'''
# ccc doesn't look at meta_param_dict for validating inputs.
params = Specifications()
params.adjust(adjustment["ccc"], raise_errors=False)
errors_warnings["ccc"]["errors"].update(params.errors)
return errors_warnings
def test_implement_bad_reform1():
specs = Specifications()
# profit rate has an upper bound at 1.0
new_specs = {'profit_rate': 1.2}
specs.update_specifications(new_specs, raise_errors=False)
assert len(specs.parameter_errors) > 0
assert specs.parameter_errors == 'ERROR: profit_rate value 1.2 > max value 1.0\n'
assert len(specs.parameter_warnings) == 0
def test_implement_bad_reform2():
specs = Specifications()
# Pick a category for depreciation that is out of bounds
new_specs = {'profit_rate': 0.5, 'DeprecSystem_3yr': 'not_a_deprec_system'}
specs.update_specifications(new_specs, raise_errors=False)
assert len(specs.parameter_errors) > 0
assert specs.parameter_errors == "ERROR: DeprecSystem_3yr value ['not_a_deprec_system'] not in possible values ['GDS', 'ADS', 'Economic']\n"
assert len(specs.parameter_warnings) == 0
def test_implement_reform():
specs = Specifications()
new_specs = {'profit_rate': 0.4, 'm': 0.5, 'start_year': 2019}
specs.update_specifications(new_specs)
assert specs.profit_rate == 0.4
assert specs.m == 0.5
assert specs.start_year == 2019
assert len(specs.parameter_errors) == 0
assert len(specs.parameter_warnings) == 0
def test_implement_bad_reform1():
specs = Specifications()
# profit rate has an upper bound at 1.0
new_specs = {'profit_rate': 1.2}
specs.update_specifications(new_specs, raise_errors=False)
assert len(specs.errors) > 0
print(specs.errors)
exp = {'profit_rate': ['profit_rate 1.2 must be less than 1.0.']}
assert specs.errors == exp
def test_bubble_widget():
'''
Test asset bubble plot method.
'''
assets = Assets()
p = Specifications()
calc = Calculator(p, assets)
p2 = Specifications(year=2026)
p2.update_specifications({'CIT_rate': 0.25})
calc2 = Calculator(p2, assets)
fig = calc.bubble_widget(calc2)
assert fig
def test_implement_bad_reform1():
specs = Specifications()
# profit rate has an upper bound at 1.0
new_specs = {
'profit_rate': 1.2
}
specs.update_specifications(new_specs, raise_errors=False)
assert len(specs.parameter_errors) > 0
assert specs.parameter_errors == 'ERROR: profit_rate value 1.2 > max value 1.0\n'
assert len(specs.parameter_warnings) == 0
def test_implement_bad_reform2():
specs = Specifications()
# Pick a category for depreciation that is out of bounds
new_specs = {
'profit_rate': 0.5,
'DeprecSystem_3yr': 'not_a_deprec_system'
}
specs.update_specifications(new_specs, raise_errors=False)
assert len(specs.parameter_errors) > 0
assert specs.parameter_errors == "ERROR: DeprecSystem_3yr value ['not_a_deprec_system'] not in possible values ['GDS', 'ADS', 'Economic']\n"
assert len(specs.parameter_warnings) == 0
def get_inputs(meta_params_dict):
'''
Function to get user input parameters from COMP
'''
meta_params = MetaParams()
meta_params.adjust(meta_params_dict)
params = Specifications()
spec = params.specification(meta_data=True,
serializable=True,
year=meta_params.year)
return (meta_params.specification(meta_data=True, serializable=True), {
"ccc": spec
})
def test_implement_reform():
specs = Specifications()
new_specs = {
'profit_rate': 0.4,
'm': 0.5,
'start_year': 2019
}
specs.update_specifications(new_specs)
assert specs.profit_rate == 0.4
assert specs.m == 0.5
assert specs.start_year == 2019
assert len(specs.parameter_errors) == 0
assert len(specs.parameter_warnings) == 0
def test_implement_bad_reform2():
specs = Specifications()
# Pick a category for depreciation that is out of bounds
new_specs = {'profit_rate': 0.5, 'DeprecSystem_3yr': 'not_a_deprec_system'}
specs.update_specifications(new_specs, raise_errors=False)
assert len(specs.errors) > 0
exp = {
'DeprecSystem_3yr': [
'DeprecSystem_3yr "not_a_deprec_system" must be in list of choices GDS, ADS, Economic.'
]
}
assert specs.errors == exp
def test_asset_bubble():
'''
Test asset bubble plot method.
'''
assets = Assets()
p = Specifications()
calc = Calculator(p, assets)
p2 = Specifications(year=2026)
p2.update_specifications({'CIT_rate': 0.25})
calc2 = Calculator(p2, assets)
fig = calc.asset_bubble(calc2)
assert fig
fig = calc.asset_bubble(calc2, output_variable='rho_mix')
assert fig
def test_grouped_bar():
'''
Test grouped_bar method.
'''
assets = Assets()
p = Specifications()
calc = Calculator(p, assets)
p2 = Specifications(year=2026)
p2.update_specifications({'CIT_rate': 0.25})
calc2 = Calculator(p2, assets)
fig = calc.grouped_bar(calc2)
assert fig
fig = calc.grouped_bar(calc2, output_variable='rho', group_by_asset=False)
assert fig
def test_range_plot():
'''
Test range_plot method.
'''
assets = Assets()
p = Specifications()
calc = Calculator(p, assets)
p2 = Specifications(year=2026)
p2.update_specifications({'CIT_rate': 0.25})
calc2 = Calculator(p2, assets)
fig = calc.range_plot(calc2)
assert fig
fig = calc.range_plot(calc2, output_variable='rho')
assert fig
def test_summary_table():
'''
Test difference_table method.
'''
yr = 2018
assets = Assets()
p = Specifications(year=yr)
calc1 = Calculator(p, assets)
assert calc1.current_year == yr
reform = {'CIT_rate': 0.38}
p.update_specifications(reform)
calc2 = Calculator(p, assets)
assert calc2.current_year == yr
summary_df = calc1.summary_table(calc2)
assert isinstance(summary_df, pd.DataFrame)
def test_params_start_year(year, expected_values):
'''
Test that different start years return the expected parameter values
as specificed in the default_parameters.json file.
'''
p = Specifications(year=year)
assert (np.allclose(p.u['c'], expected_values[0]))
assert (np.allclose(p.bonus_deprec['3'], expected_values[1]))
assert (np.allclose(p.phi, expected_values[2]))
def test_calc_by_industry():
'''
Test difference_table method.
'''
yr = 2018
assets = Assets()
p = Specifications(year=yr)
calc1 = Calculator(p, assets)
assert calc1.current_year == yr
asset_df = calc1.calc_by_industry()
assert isinstance(asset_df, pd.DataFrame)
def run_model(meta_param_dict, adjustment):
'''
Initiliazes classes from CCC that compute the model under
different policies. Then calls function get output objects.
'''
meta_params = MetaParams()
meta_params.adjust(meta_param_dict)
if meta_params.data_source == "PUF":
data = retrieve_puf(AWS_ACCESS_KEY_ID, AWS_SECRET_ACCESS_KEY)
else:
data = "cps"
params = Specifications(year=meta_params.year, call_tc=True, data=data)
params.adjust(adjustment["ccc"])
assets = Assets()
calc1 = Calculator(params, assets)
params2 = Specifications(year=meta_params.year)
calc2 = Calculator(params2, assets)
comp_dict = comp_output(calc1, calc2)
return comp_dict
from ccc.calculator import Calculator
from bokeh.io import show
# Read in a reform to compare against
# Note that TCJA is current law baseline in TC 0.16+
# Thus to compare TCJA to 2017 law, we'll use 2017 law as the reform
reform_url = ('https://raw.githubusercontent.com/'
'PSLmodels/Tax-Calculator/master/taxcalc/'
'reforms/2017_law.json')
ref = taxcalc.Calculator.read_json_param_objects(reform_url, None)
iit_reform = ref['policy']
# Initialize Asset and Calculator Objects
assets = Assets()
# Baseline
baseline_parameters = Specifications(year=2019, call_tc=False, iit_reform={})
calc1 = Calculator(baseline_parameters, assets)
# Reform
reform_parameters = Specifications(year=2019, call_tc=False, iit_reform={})
business_tax_adjustments = {
'CIT_rate': 0.35,
'BonusDeprec_3yr': 0.50,
'BonusDeprec_5yr': 0.50,
'BonusDeprec_7yr': 0.50,
'BonusDeprec_10yr': 0.50,
'BonusDeprec_15yr': 0.50,
'BonusDeprec_20yr': 0.50
}
reform_parameters.update_specifications(business_tax_adjustments)
calc2 = Calculator(reform_parameters, assets)
# Read in a reform to compare against
# Note that TCJA is current law baseline in TC 0.16+
# Thus to compare TCJA to 2017 law, we'll use 2017 law as the reform
reform_url = ('https://raw.githubusercontent.com/'
'PSLmodels/Tax-Calculator/master/taxcalc/'
'reforms/2017_law.json')
ref = taxcalc.Calculator.read_json_param_objects(reform_url, None)
iit_reform = ref['policy']
# Initialize Asset and Calculator Objects
assets = Assets()
# Baseline
baseline_parameters = Specifications(year=2018, call_tc=True, iit_reform={})
calc1 = Calculator(baseline_parameters, assets)
# Reform
reform_parameters = Specifications(year=2018, call_tc=True,
iit_reform=iit_reform)
business_tax_adjustments = {
'CIT_rate': 0.35, 'BonusDeprec_3yr': 0.50, 'BonusDeprec_5yr': 0.50,
'BonusDeprec_7yr': 0.50, 'BonusDeprec_10yr': 0.50,
'BonusDeprec_15yr': 0.50, 'BonusDeprec_20yr': 0.50}
reform_parameters.update_specifications(business_tax_adjustments)
calc2 = Calculator(reform_parameters, assets)
# Do calculations by asset
base_assets_df = calc1.calc_by_asset()
reform_assets_df = calc2.calc_by_asset()
# Do Calculations by Industry
base_industry_df = calc1.calc_by_industry()
reform_industry_df = calc2.calc_by_industry()
# Generate dataframes with differences
import numpy as np
from pandas.util.testing import assert_frame_equal
from ccc import run_ccc
from ccc.parameters import Specifications
CUR_PATH = os.path.abspath(os.path.dirname(__file__))
# Load input values
input_tuple = tuple(json.load(open(os.path.join(
CUR_PATH, 'run_ccc_inputs.json'))))
test_run, baseline, start_year, iit_reform, data, user_params = input_tuple
result_by_asset = pd.read_json(os.path.join(
CUR_PATH, 'run_ccc_asset_output.json'))
result_by_industry = pd.read_json(os.path.join(
CUR_PATH, 'run_ccc_industry_output.json'))
parameters = Specifications(year=start_year, call_tc=True,
iit_reform=iit_reform)
test_by_asset, test_by_industry = run_ccc.run_ccc(
parameters, baseline, data='cps')
# Lists of variables to compare
asset_var_list = ['delta', 'z_c', 'z_c_d', 'z_c_e', 'z_nc',
'z_nc_d', 'z_nc_e', 'rho_c', 'rho_c_d', 'rho_c_e',
'rho_nc', 'rho_nc_d', 'rho_nc_e', 'ucc_c',
'ucc_c_d', 'ucc_c_e', 'ucc_nc', 'ucc_nc_d',
'ucc_nc_e', 'metr_c', 'metr_c_d', 'metr_c_e',
'metr_nc', 'metr_nc_d', 'metr_nc_e', 'mettr_c',
'mettr_c_d', 'mettr_c_e', 'mettr_nc',
'mettr_nc_d', 'mettr_nc_e', 'assets_c',
'assets_nc']
industry_var_list = ['delta_c', 'delta_c', 'z_c', 'z_c_d', 'z_c_e',
'z_nc', 'z_nc_d', 'z_nc_e', 'rho_c', 'rho_c_d',
def test_read_json_params_objects(revision_file):
exp = {"revision": {"CIT_rate": 0.3}}
act1 = Specifications.read_json_param_objects(JSON_REVISION_FILE)
assert exp == act1
act2 = Specifications.read_json_param_objects(JSON_REVISION_FILE)
assert exp == act2
import pandas as pd
import numpy as np
import pytest
from pandas.testing import assert_series_equal, assert_frame_equal
from ccc import calcfunctions as cf
from ccc.parameters import Specifications
p = Specifications()
df = pd.DataFrame.from_dict({
'Method': [
'DB 200%', 'DB 150%', 'SL', 'Economic', 'Expensing', 'DB 200%',
'DB 150%', 'SL', 'Economic', 'Expensing'
],
'GDS Life': [10, 10, 3, 15, 27.5, 27.5, 10, 3, 15, 7],
'ADS Life': [10, 10, 3, 15, 27.5, 27.5, 10, 3, 15, 7],
'GDS Class Life': [10, 10, 3, 15, 27.5, 27.5, 10, 3, 15, 7]
})
expected_df = df.copy()
expected_df['System'] = pd.Series(
['GDS', 'GDS', 'GDS', 'GDS', 'GDS', 'GDS', 'GDS', 'GDS', 'GDS', 'GDS'],
index=expected_df.index)
expected_df['bonus'] = pd.Series(
[1.0, 1.0, 1.0, 1.0, 0.0, 0.0, 1.0, 1.0, 1.0, 1.0],
index=expected_df.index)
expected_df['b'] = pd.Series([2, 1.5, 1, 1, 1, 2, 1.5, 1, 1, 1],
index=expected_df.index)
expected_df['Y'] = pd.Series([10, 10, 3, 15, 27.5, 27.5, 10, 3, 15, 7],
index=expected_df.index)
test_data = [(df, p, expected_df)]
from ccc.run_ccc import run_ccc_with_baseline_delta
from taxcalc import *
from ccc.parameters import Specifications
test_run = False # flag for test run (for Travis CI)
start_year = 2018
# Note that TCJA is current law baseline in TC 0.16+
# Thus to compare TCJA to 2017 law, we'll use 2017 law as the reform
reform_url = ('https://raw.githubusercontent.com/'
'PSLmodels/Tax-Calculator/master/taxcalc/'
'reforms/2017_law.json')
ref = Calculator.read_json_param_objects(reform_url, None)
iit_reform = ref['policy']
# Initialize Cost-of-Capital-Calculator parameters
baseline_parameters = Specifications(year=2018, call_tc=True,
iit_reform={})
reform_parameters = Specifications(year=2018, call_tc=True,
iit_reform=iit_reform)
reform_params = {'CIT_rate': 0.35, 'BonusDeprec_3yr': 0.50,
'BonusDeprec_5yr': 0.50, 'BonusDeprec_7yr': 0.50,
'BonusDeprec_10yr': 0.50, 'BonusDeprec_15yr': 0.50,
'BonusDeprec_20yr': 0.50}
reform_parameters.update_specifications(reform_params)
# Run Cost-of-Capital-Calculator
run_ccc_with_baseline_delta(baseline_parameters, reform_parameters,
data='cps')
def test_create_specs_object():
specs = Specifications()
assert specs
def test_read_json_params_objects(revision_file):
exp = {"revision": {"CIT_rate": 0.3}}
act1 = Specifications.read_json_param_objects(JSON_REVISION_FILE)
assert exp == act1
act2 = Specifications.read_json_param_objects(JSON_REVISION_FILE)
assert exp == act2