Merge pull request #930 from Mv77/Calibration/cpi

Module for inflation adjustments based on CPI
econ-ark · Jan 27, 2021 · dd2f9cb · dd2f9cb
2 parents 8bc1103 + 7bf7ae2
commit dd2f9cb
Show file tree

Hide file tree

Showing 7 changed files with 231 additions and 2 deletions.
diff --git a/Documentation/CHANGELOG.md b/Documentation/CHANGELOG.md
@@ -20,9 +20,10 @@ Release Data: TBD
 * Centralizes the definition of value, marginal value, and marginal marginal value functions that use inverse-space
 interpolation for problems with CRRA utility. See [#888](https://github.com/econ-ark/HARK/pull/888).
 * MarkovProcess class used in ConsMarkovModel, ConsRepAgentModel, ConsAggShockModel [#902](https://github.com/econ-ark/HARK/pull/902) [#929](https://github.com/econ-ark/HARK/pull/929)
+* Adds SSA life tables and methods to extract survival probabilities from them [#986](https://github.com/econ-ark/HARK/pull/906).
+* Adds the U.S. CPI research series and tools to extract inflation adjustments from it [#930](https://github.com/econ-ark/HARK/pull/930).
 * replace HARKobject base class with MetricObject and Model classes [#903](https://github.com/econ-ark/HARK/pull/903/)
 * Add __repr__ and __eq__ methods to Model class [#903](https://github.com/econ-ark/HARK/pull/903/)
-* Adds a SSA life tables and methods to extract survival probabilities from them [#986](https://github.com/econ-ark/HARK/pull/906).
 * Fix the return fields of `dcegm/calcCrossPoints`[#909](https://github.com/econ-ark/HARK/pull/909).
 * Corrects location of constructor documentation to class string for Sphinx rendering [#908](https://github.com/econ-ark/HARK/pull/908)
 

diff --git a/HARK/datasets/cpi/__init__.py b/HARK/datasets/cpi/__init__.py
diff --git a/HARK/datasets/cpi/us/CPITools.py b/HARK/datasets/cpi/us/CPITools.py
@@ -0,0 +1,134 @@
+# -*- coding: utf-8 -*-
+"""
+Created on Wed Jan 20 18:07:41 2021
+
+@author: Mateo
+"""
+
+import os
+import urllib.request
+import pandas as pd
+import warnings
+import numpy as np
+
+__all__ = ["get_cpi_series", "cpi_deflator"]
+
+us_cpi_dir = os.path.dirname(os.path.abspath(__file__))
+
+
+def download_cpi_series():
+    """
+    A method that downloads the cpi research series file directly from the
+    bls site onto the working directory.
+    After being converted to a .csv, this is the file that the rest of
+    the functions in this script use and must be placed in HARK/datasets/cpi/us.
+    This function is not for users but for whenever mantainers want to update
+    the cpi series as new data comes out.
+
+    Returns
+    -------
+    None.
+
+    """
+    urllib.request.urlretrieve(
+        "https://www.bls.gov/cpi/research-series/r-cpi-u-rs-allitems.xlsx",
+        "r-cpi-u-rs-allitems.xlsx",
+    )
+
+
+def get_cpi_series():
+    """
+    This function reads the cpi series currently in the toolbox and returns it
+    as a pandas dataframe.
+
+    Returns
+    -------
+    cpi : Pandas DataFrame
+        DataFrame representation of the CPI research series file from the
+        Bureau of Labor Statistics.
+
+    """
+
+    cpi = pd.read_csv(
+        os.path.join(us_cpi_dir, "r-cpi-u-rs-allitems.csv"),
+        skiprows=5,
+        index_col=0,
+    )
+    return cpi
+
+
+def cpi_deflator(from_year, to_year, base_month=None):
+    """
+    Finds cpi deflator to transform quantities measured in "from_year" U.S.
+    dollars to "to_year" U.S. dollars.
+    The deflators are computed using the "r-cpi-u-rs" series from the BLS.
+
+    Parameters
+    ----------
+    from_year : int
+        Base year in which the nominal quantities are currently expressed.
+    to_year : int
+        Target year in which you wish to express the quantities.
+    base_month : str, optional
+        Month at which to take the CPI measurements to calculate the deflator.
+        The default is None, and in this case annual averages of the CPI are
+        used.
+
+    Returns
+    -------
+    deflator : numpy array
+        A length-1 numpy array with the deflator that, when multiplied by the
+        original nominal quantities, rebases them to "to_year" U.S. dollars.
+
+    """
+
+    # Check years are conforming
+    assert type(from_year) is int and type(to_year) is int, "Years must be integers."
+
+    # Check month is conforming
+    if base_month is not None:
+
+        months = [
+            "JAN",
+            "FEB",
+            "MAR",
+            "APR",
+            "MAY",
+            "JUNE",
+            "JULY",
+            "AUG",
+            "SEP",
+            "OCT",
+            "NOV",
+            "DEC",
+        ]
+
+        assert base_month in months, (
+            "If a month is provided, it must be " + "one of " + ",".join(months) + "."
+        )
+
+        column = base_month
+
+    else:
+
+        warnings.warn("No base month was provided. Using annual CPI averages.")
+        column = "AVG"
+
+    # Get cpi and subset the columns we need.
+    cpi = get_cpi_series()
+    cpi_series = cpi[[column]].dropna()
+
+    try:
+
+        deflator = np.divide(
+            cpi_series.loc[to_year].to_numpy(), cpi_series.loc[from_year].to_numpy()
+        )
+
+    except KeyError as e:
+
+        message = (
+            "Could not find a CPI value for the requested " + "year-month combinations."
+        )
+        raise Exception(message).with_traceback(e.__traceback__)
+
+    return deflator
diff --git a/HARK/datasets/cpi/us/README.md b/HARK/datasets/cpi/us/README.md
@@ -0,0 +1,22 @@
+# Inflation adjustments using the CPI research series.
+
+This folder contains tools for transforming nominal U.S. dollar quantities to different base years using the consumer price index.
+
+## Data
+
+The dataset is stored in file `./r-cpi-u-rs-allitems.csv`, which comes from the [U.S. Bureau of Labor Statistics](https://www.bls.gov/cpi/research-series/r-cpi-u-rs-home.htm).
+As of January 21, 2021 the direct link to the file was [https://www.bls.gov/cpi/research-series/r-cpi-u-rs-allitems.xlsx](https://www.bls.gov/cpi/research-series/r-cpi-u-rs-allitems.xlsx). The BLS originally distributes the series as a `.xlsx` file. The file found here is a copy of the original, converted to `.csv`.
+
+The file contains the monthly research series (retroactive, using current methods) of the consumer price index (all items) starting in 1977.
+
+### Format
+
+The file is formatted so that each row represents a year, with CPI measurements for each month stored in columns in ascending order. There is an
+additional column labeled `AVG` which corresponds to the average of CPI measurements across months.
+
+## Tools
+
+`CPITools.py` contains functions to:
+- Download the latest series file from the BLS to the working directory.
+- Read and format the CPI series as a Pandas DataFrame.
+- Produce CPI deflators using the series for any given pair of years (between 1977 and 2019) and taking any month as base.
diff --git a/HARK/datasets/cpi/us/__init__.py b/HARK/datasets/cpi/us/__init__.py
diff --git a/HARK/datasets/cpi/us/r-cpi-u-rs-allitems.csv b/HARK/datasets/cpi/us/r-cpi-u-rs-allitems.csv
@@ -0,0 +1,49 @@
+Consumer Price Index Research Series Using Current Methods (CPI-U-RS),,,,,,,,,,,,,
+U.S. city average,,,,,,,,,,,,,
+All items,,,,,,,,,,,,,
+Not Seasonally Adjusted,,,,,,,,,,,,,
+December 1977=100,,,,,,,,,,,,,
+YEAR,JAN,FEB,MAR,APR,MAY,JUNE,JULY,AUG,SEP,OCT,NOV,DEC,AVG
+1977,,,,,,,,,,,,100,
+1978,100.5,101.1,101.8,102.7,103.6,104.5,105,105.5,106.1,106.7,107.3,107.8,104.4
+1979,108.7,109.7,110.7,111.8,113,114.1,115.1,116,117.1,117.9,118.5,119.5,114.3
+1980,120.8,122.4,123.8,124.7,125.7,126.7,127.5,128.6,129.9,130.7,131.5,132.4,127.1
+1981,133.6,135.2,136.3,137.1,137.9,138.7,139.7,140.7,141.8,142.4,142.9,143.4,139.1
+1982,144.2,144.7,144.9,145,146.1,147.5,148.5,148.8,149.5,150.2,150.5,150.6,147.5
+1983,151,151.1,151.2,152.4,153.2,153.7,154.3,154.8,155.6,156,156.2,156.4,153.8
+1984,157.2,158,158.3,159.1,159.5,160,160.5,161.1,161.8,162.2,162.2,162.4,160.2
+1985,162.6,163.2,164,164.6,165.3,165.7,166,166.4,166.9,167.3,167.9,168.2,165.7
+1986,168.7,168.3,167.5,167.1,167.6,168.4,168.4,168.7,169.6,169.7,169.7,169.9,168.6
+1987,170.9,171.5,172.3,173.1,173.6,174.3,174.6,175.5,176.4,176.7,176.9,176.7,174.4
+1988,177.2,177.5,178.3,179.1,179.7,180.4,181.1,181.8,183,183.4,183.6,183.7,180.7
+1989,184.5,185.2,186.2,187.4,188.4,188.8,189.3,189.4,190.1,190.9,191.2,191.4,188.6
+1990,193.3,194.1,195.2,195.5,195.8,196.8,197.6,199.3,200.9,202,202.3,202.3,197.9
+1991,203.2,203.4,203.6,203.8,204.4,204.8,205,205.5,206.4,206.5,207,207.1,205.1
+1992,207.5,208.1,208.9,209.3,209.6,210.1,210.4,210.9,211.5,212.1,212.4,212.3,210.2
+1993,212.9,213.6,214.3,214.9,215.3,215.5,215.6,216.1,216.3,217.1,217.2,217,215.5
+1994,217.4,218,218.7,219,219.2,219.8,220.3,221.1,221.4,221.6,221.8,221.7,220
+1995,222.5,223.2,223.9,224.6,225,225.5,225.6,226,226.5,227,226.9,226.7,225.3
+1996,227.8,228.6,229.8,230.6,231.1,231.3,231.6,231.9,232.7,233.3,233.7,233.7,231.3
+1997,234.4,235.1,235.6,235.8,235.8,236.1,236.2,236.7,237.4,237.8,237.8,237.3,236.3
+1998,237.8,238.1,238.5,239,239.3,239.5,239.7,240.1,240.4,240.9,240.8,240.6,239.5
+1999,241.3,241.6,242.3,244,244,244.1,244.7,245.4,246.5,247.1,247.2,247.2,244.6
+2000,248,249.4,251.4,251.6,251.8,253.2,253.7,253.8,255.1,255.5,255.7,255.5,252.9
+2001,257.2,258.2,258.9,259.8,260.9,261.5,260.7,260.8,261.8,261,260.5,259.5,260.1
+2002,260.2,261.2,262.6,264.1,264,264.2,264.5,265.4,265.8,266.3,266.3,265.7,264.2
+2003,266.9,269,270.6,269.9,269.5,269.9,270.2,271.2,272,271.8,271,270.7,270.2
+2004,272.1,273.6,275.3,276.2,277.7,278.6,278.3,278.4,278.9,280.4,280.6,279.6,277.5
+2005,280.1,281.7,283.9,285.8,285.6,285.7,287,288.5,292,292.6,290.3,289,286.9
+2006,291.3,291.9,293.6,296,297.4,298.1,298.9,299.6,298.1,296.5,296,296.5,296.2
+2007,297.4,299,301.7,303.6,305.5,306.1,306,305.5,306.3,306.9,308.8,308.6,304.6
+2008,310.1,311,313.7,315.6,318.2,321.5,323.2,321.9,321.4,318.2,312.1,308.8,316.3
+2009,310.2,311.7,312.5,313.3,314.2,316.9,316.4,317.1,317.3,317.6,317.8,317.3,315.2
+2010,318.3,318.4,319.7,320.3,320.5,320.2,320.3,320.7,320.9,321.3,321.5,322,320.4
+2011,323.6,325.2,328.3,330.5,332,331.7,332,332.9,333.4,332.7,332.4,331.6,330.5
+2012,333.1,334.6,337.1,338.1,337.7,337.3,336.7,338.6,340.1,340,338.4,337.5,337.5
+2013,338.6,341.4,342.2,341.9,342.5,343.3,343.5,343.9,344.3,343.4,342.8,342.8,342.5
+2014,344.1,345.4,347.6,348.7,350,350.7,350.5,349.9,350.2,349.3,347.5,345.6,348.3
+2015,344,345.5,347.5,348.3,350,351.3,351.3,350.8,350.3,350.2,349.4,348.3,348.9
+2016,348.8,349.1,350.6,352.3,353.7,354.9,354.3,354.7,355.5,356,355.4,355.5,353.4
+2017,357.6,358.7,359,360.1,360.4,360.7,360.5,361.6,363.5,363.2,363.3,363.1,361
+2018,365,366.7,367.5,369,370.5,371.1,371.1,371.3,371.8,372.4,371.2,370,369.8
+2019,370.7,372.2,374.3,376.3,377.1,377.2,377.8,377.8,378.1,379,378.8,378.5,376.5
diff --git a/HARK/datasets/tests/test_load_data.py b/HARK/datasets/tests/test_load_data.py
@@ -1,6 +1,6 @@
 import unittest
 from HARK.datasets import load_SCF_wealth_weights
-
+from HARK.datasets.cpi.us.CPITools import cpi_deflator
 
 class test_load_SCF_wealth_weights(unittest.TestCase):
     def setUp(self):
@@ -9,3 +9,26 @@ def setUp(self):
     def test_shape(self):
         self.assertEqual(self.SCF_wealth.shape, (3553,))
         self.assertEqual(self.SCF_weights.shape, (3553,))
+
+# %% US CPI tests
+class test_cpi_deflators(unittest.TestCase):
+
+    def test_month_deflators(self):
+
+        # Same year test
+        defl_same_year = cpi_deflator(2000, 2000, 'SEP')
+        self.assertEqual(defl_same_year[0], 1.0)
+
+        # Different year test
+        defl_diff_year = cpi_deflator(1998, 2019, 'SEP')
+        self.assertAlmostEqual(defl_diff_year[0], 1.57279534)
+
+    def test_avg_deflators(self):
+
+        # Same year test
+        defl_same_year = cpi_deflator(2000, 2000)
+        self.assertEqual(defl_same_year[0], 1.0)
+
+        # Different year test
+        defl_diff_year = cpi_deflator(1998, 2019)
+        self.assertAlmostEqual(defl_diff_year[0], 1.57202505)