From 4054ebf3ebfc3fa855fba469fa6a6fd39c2b744f Mon Sep 17 00:00:00 2001
From: martinholmer <martin.holmer@gmail.com>
Date: Sun, 4 Mar 2018 07:38:40 -0500
Subject: [PATCH] Revise Calculator.decile_graph and associated utility
 functions

---
 taxcalc/calculate.py        | 18 ++++++++++++++----
 taxcalc/tests/test_utils.py |  8 ++++----
 taxcalc/utils.py            | 14 ++++++++++++--
 3 files changed, 30 insertions(+), 10 deletions(-)

diff --git a/taxcalc/calculate.py b/taxcalc/calculate.py
index c1bc9db9c..14f34ee48 100644
--- a/taxcalc/calculate.py
+++ b/taxcalc/calculate.py
@@ -990,7 +990,7 @@ def pch_graph(self, calc):
                              title='')
         return fig
 
-    def decile_graph(self, calc):
+    def decile_graph(self, calc, hide_negative_incomes=True):
         """
         Create graph that shows percentage change in aftertax expanded
         income (from going from policy in self to policy in calc) for
@@ -1000,8 +1000,6 @@ def decile_graph(self, calc):
         immediately in an interactive or notebook session (following
         the instructions in the documentation of the xtr_graph_plot
         utility function).
-        Note that the bottom decile result omits filing units with
-        negative or zero baseline (self) expanded income.
 
         Parameters
         ----------
@@ -1010,6 +1008,17 @@ def decile_graph(self, calc):
             where both self and calc have calculated taxes for this year
             before being used by this method
 
+        hide_negative_incomes : boolean
+            if True (which is the default), the bottom table bin containing
+            filing units with non-positive expanded_income is not shown in
+            the graph and the table bin containing filing units with positive
+            expanded_income in the bottom decile is shown with its bar width
+            adjusted to the number of weighted filing units in bottom decile
+            who have positive expanded_income; if False, the bottom table bin
+            containing filing units with non-positive expanded_income is shown,
+            which may be misleading because the percentage change is correctly
+            calculated with a negative divisor.
+
         Returns
         -------
         graph that is a bokeh.plotting figure object
@@ -1023,7 +1032,8 @@ def decile_graph(self, calc):
                                            income_measure='expanded_income',
                                            tax_to_diff='combined')
         # construct data for graph
-        data = dec_graph_data(diff_table, year=self.current_year)
+        data = dec_graph_data(diff_table, year=self.current_year,
+                              hide_negative_incomes=hide_negative_incomes)
         # construct figure from data
         fig = dec_graph_plot(data,
                              width=850,
diff --git a/taxcalc/tests/test_utils.py b/taxcalc/tests/test_utils.py
index 93268558c..e362847e7 100644
--- a/taxcalc/tests/test_utils.py
+++ b/taxcalc/tests/test_utils.py
@@ -988,10 +988,10 @@ def test_dec_qin_graph_plots(cps_subsample):
     assert calc1.current_year == calc2.current_year
     calc1.calc_all()
     calc2.calc_all()
-    fig_dec = calc1.decile_graph(calc2)
-    assert fig_dec
-    fig_qin = calc1.quintile_graph(calc2)
-    assert fig_qin
+    fig_dec_hide = calc1.decile_graph(calc2)
+    assert fig_dec_hide
+    fig_dec_show = calc1.decile_graph(calc2, hide_negative_incomes=False)
+    assert fig_dec_show
 
 
 def test_nonsmall_diffs():
diff --git a/taxcalc/utils.py b/taxcalc/utils.py
index 638859953..4ba7ea2db 100644
--- a/taxcalc/utils.py
+++ b/taxcalc/utils.py
@@ -1445,7 +1445,7 @@ def bootstrap_se_ci(data, seed, num_samples, statistic, alpha):
     return bsest
 
 
-def dec_graph_data(diff_table, year):
+def dec_graph_data(diff_table, year, hide_negative_incomes=True):
     """
     Prepare data needed by dec_graph_plot utility function.
 
@@ -1464,7 +1464,14 @@ def dec_graph_data(diff_table, year):
     # construct dictionary containing the bar data required by dec_graph_plot
     bars = dict()
     nbins = len(DECILE_ROW_NAMES)
-    for idx in range(0, nbins):
+    if hide_negative_incomes:
+        first_bin = 1
+        bottom_size = (diff_table['count'][1] /
+                       (diff_table['count'][0] + diff_table['count'][1]))
+    else:
+        first_bin = 0
+        bottom_size = 1.0
+    for idx in range(first_bin, nbins):
         info = dict()
         info['label'] = DECILE_ROW_NAMES[idx]
         info['value'] = diff_table['pc_aftertaxinc'][idx]
@@ -1475,6 +1482,7 @@ def dec_graph_data(diff_table, year):
     # construct dictionary containing bar data and auto-generated labels
     data = dict()
     data['bars'] = bars
+    data['bottom_bar_size'] = bottom_size
     xlabel = 'Reform-Induced Percentage Change in After-Tax Expanded Income'
     data['xlabel'] = xlabel
     ylabel = 'Expanded Income Percentile Group'
@@ -1576,6 +1584,8 @@ def dec_graph_plot(data,
         bval = data['bars'][idx]['value']
         blabel = data['bars'][idx]['label']
         bheight = barheight
+        if yidx == 0:
+            bheight *= data['bottom_bar_size']
         if blabel == '90-95':
             bheight *= 0.5
             bcolor = 'red'