taxcalc 5.3.0__py3-none-any.whl → 6.0.0__py3-none-any.whl

taxcalc/taxcalcio.py

@@ -9,6 +9,7 @@ import os
 import gc
 import copy
 import sqlite3
+from pathlib import Path
 import numpy as np
 import pandas as pd
 import paramtools
@@ -75,11 +76,9 @@ class TaxCalcIO():
         self.errmsg = ''
         # check name and existence of INPUT file
         inp = 'x'
-        self.puf_input_data = False
         self.cps_input_data = False
+        self.puf_input_data = False
         self.tmd_input_data = False
-        self.tmd_weights = None
-        self.tmd_gfactor = None
         if isinstance(input_data, str):
             # remove any leading directory path from INPUT filename
             fname = os.path.basename(input_data)
@@ -90,12 +89,23 @@ class TaxCalcIO():
                 msg = 'INPUT file name does not end in .csv'
                 self.errmsg += f'ERROR: {msg}\n'
             # check existence of INPUT file
-            self.puf_input_data = input_data.endswith('puf.csv')
             self.cps_input_data = input_data.endswith('cps.csv')
+            self.puf_input_data = input_data.endswith('puf.csv')
             self.tmd_input_data = input_data.endswith('tmd.csv')
             if not self.cps_input_data and not os.path.isfile(input_data):
                 msg = 'INPUT file could not be found'
                 self.errmsg += f'ERROR: {msg}\n'
+            # if puf_input_data is True, construct weights and ratios paths
+            if self.puf_input_data:  # pragma: no cover
+                puf_dir = os.path.dirname(input_data)
+                self.puf_weights = os.path.join(puf_dir, 'puf_weights.csv.gz')
+                self.puf_ratios = os.path.join(puf_dir, 'puf_ratios.csv')
+                if not os.path.isfile(self.puf_weights):
+                    msg = f'weights file {self.puf_weights} could not be found'
+                    self.errmsg += f'ERROR: {msg}\n'
+                if not os.path.isfile(self.puf_ratios):
+                    msg = f'gfactor file {self.puf_ratios} could not be found'
+                    self.errmsg += f'ERROR: {msg}\n'
             # if tmd_input_data is True, construct weights and gfactor paths
             if self.tmd_input_data:  # pragma: no cover
                 tmd_dir = os.path.dirname(input_data)
@@ -253,7 +263,7 @@ class TaxCalcIO():
             delete_file(self.output_filename.replace('.xxx', ext))
 
     def init(self, input_data, tax_year, baseline, reform, assump,
-             aging_input_data, exact_calculations):
+             exact_calculations):
         """
         TaxCalcIO class post-constructor method that completes initialization.
 
@@ -262,10 +272,6 @@ class TaxCalcIO():
         First five are same as the first five of the TaxCalcIO constructor:
             input_data, tax_year, baseline, reform, assump.
 
-        aging_input_data: boolean
-            whether or not to extrapolate Records data from data year to
-            tax_year.
-
         exact_calculations: boolean
             specifies whether or not exact tax calculations are done without
             any smoothing of "stair-step" provisions in the tax law.
@@ -406,48 +412,46 @@ class TaxCalcIO():
         pol_ref.set_year(tax_year)
         pol_bas.set_year(tax_year)
         # read input file contents into Records objects
-        if aging_input_data:
-            if self.cps_input_data:
-                recs_ref = Records.cps_constructor(
-                    gfactors=gfactors_ref,
-                    exact_calculations=exact_calculations,
-                )
-                recs_bas = Records.cps_constructor(
-                    gfactors=gfactors_bas,
-                    exact_calculations=exact_calculations,
-                )
-            elif self.tmd_input_data:  # pragma: no cover
-                wghts = pd.read_csv(self.tmd_weights)
-                recs_ref = Records(
-                    data=pd.read_csv(input_data),
-                    start_year=Records.TMDCSV_YEAR,
-                    weights=wghts,
-                    gfactors=gfactors_ref,
-                    adjust_ratios=None,
-                    exact_calculations=exact_calculations,
-                    weights_scale=1.0,
-                )
-                recs_bas = Records(
-                    data=pd.read_csv(input_data),
-                    start_year=Records.TMDCSV_YEAR,
-                    weights=wghts,
-                    gfactors=gfactors_bas,
-                    adjust_ratios=None,
-                    exact_calculations=exact_calculations,
-                    weights_scale=1.0,
-                )
-            else:  # if not {cps|tmd}_input_data but aging_input_data: puf
-                recs_ref = Records(
-                    data=input_data,
-                    gfactors=gfactors_ref,
-                    exact_calculations=exact_calculations
-                )
-                recs_bas = Records(
-                    data=input_data,
-                    gfactors=gfactors_bas,
-                    exact_calculations=exact_calculations
-                )
+        aging_input_data = True
+        if self.cps_input_data:
+            recs_ref = Records.cps_constructor(
+                gfactors=gfactors_ref,
+                exact_calculations=exact_calculations,
+            )
+            recs_bas = Records.cps_constructor(
+                gfactors=gfactors_bas,
+                exact_calculations=exact_calculations,
+            )
+        elif self.puf_input_data:  # pragma: no cover
+            recs_ref = Records.puf_constructor(
+                data=input_data,
+                gfactors=gfactors_ref,
+                weights=self.puf_weights,
+                ratios=self.puf_ratios,
+                exact_calculations=exact_calculations,
+            )
+            recs_bas = Records.puf_constructor(
+                data=input_data,
+                gfactors=gfactors_bas,
+                weights=self.puf_weights,
+                ratios=self.puf_ratios,
+                exact_calculations=exact_calculations,
+            )
+        elif self.tmd_input_data:  # pragma: no cover
+            recs_ref = Records.tmd_constructor(
+                data_path=Path(input_data),
+                weights_path=Path(self.tmd_weights),
+                growfactors=gfactors_ref,
+                exact_calculations=exact_calculations,
+            )
+            recs_bas = Records.tmd_constructor(
+                data_path=Path(input_data),
+                weights_path=Path(self.tmd_weights),
+                growfactors=gfactors_bas,
+                exact_calculations=exact_calculations,
+            )
         else:  # input_data are raw data that are not being aged
+            aging_input_data = False
             recs_ref = Records(
                 data=input_data,
                 start_year=tax_year,
@@ -486,7 +490,7 @@ class TaxCalcIO():
         dirpath = os.path.abspath(os.path.dirname(__file__))
         return os.path.join(dirpath, self.output_filename)
 
-    def advance_to_year(self, year, aging_data):
+    def advance_to_year(self, year):
         """
         Update self.output_filename and advance Calculator objects to year.
         """
@@ -503,6 +507,11 @@ class TaxCalcIO():
         # advance baseline and reform Calculator objects to specified year
         self.calc_bas.advance_to_year(year)
         self.calc_ref.advance_to_year(year)
+        aging_data = (
+            self.cps_input_data or
+            self.puf_input_data or
+            self.tmd_input_data
+        )
         idata = 'Advance input data and' if aging_data else 'Advance'
         if not self.silent:
             print(f'{idata} policy to {year}')

taxcalc/tests/conftest.py

@@ -27,42 +27,47 @@ def tests_path_fixture():
     return os.path.abspath(os.path.dirname(__file__))
 
 
-@pytest.fixture(scope='session', name='cps_path')
-def cps_path_fixture(tests_path):
+@pytest.fixture(scope='session', name='cps_data_path')
+def cps_data_path_fixture(tests_path):
     """Fixture docstring"""
     return os.path.join(tests_path, '..', 'cps.csv.gz')
 
 
 @pytest.fixture(scope='session', name='cps_fullsample')
-def cps_fullsample_fixture(cps_path):
+def cps_fullsample_fixture(cps_data_path):
     """Fixture docstring"""
-    return pandas.read_csv(cps_path)
+    return pandas.read_csv(cps_data_path)
 
 
 @pytest.fixture(scope='session')
 def cps_subsample(cps_fullsample):
     """Fixture docstring"""
-    # draw smaller cps.csv subsample than in test_cpscsv.py
+    # draw a small cps.csv subsample
     return cps_fullsample.sample(frac=0.01, random_state=123456789)
 
 
-@pytest.fixture(scope='session', name='puf_path')
-def puf_path_fixture(tests_path):
+@pytest.fixture(scope='session', name='puf_data_path')
+def puf_data_path_fixture(tests_path):
     """Fixture docstring"""
     return os.path.join(tests_path, '..', '..', 'puf.csv')
 
 
-@pytest.fixture(scope='session', name='puf_fullsample')
-def puf_fullsample_fixture(puf_path):
+@pytest.fixture(scope='session', name='puf_weights_path')
+def puf_weights_path_fixture(tests_path):
     """Fixture docstring"""
-    return pandas.read_csv(puf_path)
+    return os.path.join(tests_path, '..', '..', 'puf_weights.csv.gz')
 
 
-@pytest.fixture(scope='session')
-def puf_subsample(puf_fullsample):
+@pytest.fixture(scope='session', name='puf_ratios_path')
+def puf_ratios_path_fixture(tests_path):
+    """Fixture docstring"""
+    return os.path.join(tests_path, '..', '..', 'puf_ratios.csv')
+
+
+@pytest.fixture(scope='session', name='tmd_data_path')
+def tmd_data_path_fixture(tests_path):
     """Fixture docstring"""
-    # draw same puf.csv subsample as in test_pufcsv.py
-    return puf_fullsample.sample(frac=0.05, random_state=2222)
+    return os.path.join(tests_path, '..', '..', 'tmd.csv')
 
 
 @pytest.fixture(scope='session', name='test_reforms_init')

taxcalc/tests/reforms_expect.csv

@@ -1,61 +1,61 @@
 rid,res1,res2,res3,res4
-1,62.1,63.1,66.2,68.9
-2,54.2,58.6,53.5,62.0
-3,75.1,76.2,80.3,84.3
+1,61.2,61.7,65.1,70.4
+2,43.5,45.9,43.1,52.8
+3,67.7,68.3,72.1,81.1
 4,0.0,0.0,0.0,0.0
 5,0.0,0.0,0.0,0.0
-6,38.4,39.2,40.0,41.0
-7,9.8,9.9,10.5,11.0
-8,1.9,2.0,2.1,2.2
-9,4.4,4.3,4.9,5.3
-10,8.2,8.6,9.1,9.5
-11,2.4,2.7,2.6,2.8
-12,0.1,0.1,0.1,0.1
-13,-33.5,-32.1,-35.0,-36.5
-14,-0.1,-0.1,-0.1,-0.1
-15,0.2,0.2,0.2,0.3
-16,-1.0,-1.0,-1.5,-1.6
+6,33.1,34.1,35.9,41.4
+7,6.8,7.0,7.7,10.7
+8,1.8,1.9,2.0,2.3
+9,5.6,5.7,5.6,7.6
+10,5.6,5.8,6.4,6.1
+11,0.0,0.0,0.0,0.0
+12,0.0,0.0,0.0,0.0
+13,-34.5,-33.0,-36.1,-38.0
+14,-0.1,-0.1,-0.1,-0.2
+15,0.5,0.4,0.3,0.5
+16,-0.7,-0.8,-1.1,-1.2
 17,-0.2,-0.2,-0.3,-0.3
-18,35.8,37.3,40.2,42.5
-19,68.2,69.9,75.1,80.2
-20,103.3,106.9,116.1,122.6
-21,68.7,70.7,76.4,80.6
-22,9.6,11.0,10.0,10.3
-23,0.3,0.3,0.3,0.4
-24,52.5,54.2,57.2,60.3
-25,2.0,2.1,2.2,2.3
-26,0.1,0.1,0.1,0.1
-27,0.2,0.2,0.2,0.2
-28,4.0,3.8,4.5,4.9
-29,15.2,13.8,17.9,19.4
-30,-6.4,-6.2,-1.6,-1.6
-31,58.2,58.6,62.5,66.3
-32,12.6,12.1,14.0,15.1
-33,7.6,7.1,8.4,9.1
-34,-1.3,-0.9,-1.5,-1.6
-35,-2.9,-2.8,-3.4,-3.5
-36,3.2,3.3,3.6,3.9
-37,30.9,32.0,35.4,37.3
-38,9.7,10.0,11.2,11.8
-39,-0.6,-0.6,-0.7,-0.7
-40,56.0,53.5,52.7,51.9
-41,0.6,0.6,0.7,0.7
+18,26.8,27.3,30.0,33.1
+19,78.5,79.4,87.3,96.4
+20,64.9,66.7,73.7,84.8
+21,39.6,40.8,45.0,52.8
+22,8.9,9.9,9.6,10.8
+23,0.0,0.0,0.0,0.0
+24,34.4,35.5,38.0,42.8
+25,1.5,1.5,1.6,1.8
+26,0.0,0.1,0.1,0.1
+27,0.1,0.1,0.1,0.2
+28,0.5,0.4,0.5,1.6
+29,22.4,20.9,25.3,27.0
+30,-3.2,-3.1,-0.8,-0.8
+31,53.7,54.2,57.4,66.2
+32,7.9,7.8,8.6,13.6
+33,3.9,3.8,4.3,6.8
+34,-0.5,-0.4,-0.8,-0.9
+35,-1.2,-1.1,-1.7,-1.9
+36,1.5,1.5,2.1,2.6
+37,15.8,16.2,18.9,25.1
+38,3.3,3.3,4.0,5.8
+39,-0.3,-0.2,-0.3,-0.4
+40,52.0,50.6,49.4,55.3
+41,0.4,0.4,0.5,0.6
 42,-0.2,-0.2,-0.2,-0.2
-43,71.8,68.8,67.6,69.4
-44,1.1,1.1,1.1,1.2
-45,-2.1,-2.0,-2.1,-2.2
-46,-2.0,-1.7,-2.0,-2.1
-47,-0.8,-0.8,-0.8,-0.7
+43,59.1,57.4,52.5,68.5
+44,1.3,1.2,1.2,1.5
+45,-2.1,-2.0,-2.1,-2.6
+46,-2.0,-1.7,-1.9,-2.4
+47,-0.9,-0.8,-0.7,-0.9
 48,-0.0,-0.0,-0.0,-0.0
-49,-27.2,-24.3,-31.9,-34.8
-50,-0.4,-0.4,-0.4,-0.4
-51,-187.0,-187.5,-193.6,-201.6
-52,-129.1,-128.5,-131.9,-136.9
-53,31.6,34.1,32.5,42.1
+49,-37.3,-36.0,-40.4,-44.7
+50,-0.1,-0.1,-0.1,-0.1
+51,-184.8,-185.3,-189.7,-206.8
+52,-129.7,-129.6,-131.1,-143.6
+53,3.5,3.9,2.9,5.6
 54,0.0,0.0,0.0,0.0
-55,-68.2,-68.4,-68.7,-68.7
-56,-68.8,-68.9,-69.2,-69.2
-57,1.4,1.7,1.7,1.4
-58,-16.8,-17.4,-23.7,-19.2
-59,-12.2,-12.2,-21.2,-13.3
-60,-46.9,-48.1,-49.3,-52.1
+55,-65.5,-71.9,-72.4,-72.5
+56,-66.6,-73.3,-73.8,-73.9
+57,-23.7,-25.7,-26.9,-28.5
+58,-25.5,-26.4,-33.6,-28.8
+59,-51.2,-52.9,-60.1,-52.3
+60,-51.7,-53.2,-54.4,-57.5