diff --git a/.Rhistory b/.Rhistory
deleted file mode 100644
index ec3efcc..0000000
--- a/.Rhistory
+++ /dev/null
@@ -1,350 +0,0 @@
-# Import necessary libraries
-import numpy as np
-import statsmodels.api as sm
-from .utils import check_inputs
-# COD, PRD, PRB functions
-def cod(ratio):
-"""
-COD is the average absolute percent deviation from the
-median ratio. It is a measure of horizontal equity in assessment.
-Horizontal equity means properties with a similar fair market value
-should be similarly assessed.
-Lower COD indicates higher uniformity/horizontal equity in assessment.
-The IAAO sets uniformity standards that define generally accepted ranges
-for COD depending on property class. See `IAAO Standard on Ratio Studies`_
-Section 9.1, Table 1.3 for a full list of standard COD ranges.
-.. _IAAO Standard on Ratio Studies: https://www.iaao.org/media/standards/Standard_on_Ratio_Studies.pdf
-.. note::
-The IAAO recommends trimming outlier ratios before calculating COD,
-as it is extremely sensitive to large outliers. The typical method used is
-dropping values beyond 3 * IQR (inner-quartile range). See
-`IAAO Standard on Ratio Studies`_ Appendix B.1.
-:param ratio:
-A numeric vector of ratios centered around 1, where the
-numerator of the ratio is the estimated fair market value and the
-denominator is the actual sale price.
-:type ratio: numeric
-:return: A numeric vector containing the COD of ``ratios``.
-:rtype: float
-:Example:
-.. code-block:: python
-# Calculate COD:
-import assesspy as ap
-ap.cod(ap.ratios_sample().ratio)
-"""
-# Input checking and error handling
-check_inputs(ratio)
-ratio = np.array(ratio)
-n = ratio.size
-median_ratio = np.median(ratio)
-cod = 100 / median_ratio * (sum(abs(ratio - median_ratio)) / n)
-return cod
-def prd(assessed, sale_price):
-"""
-PRD is the mean ratio divided by the mean ratio weighted by sale
-price. It is a measure of vertical equity in assessment. Vertical equity
-means that properties at different levels of the income distribution
-should be similarly assessed.
-PRD centers slightly above 1 and has a generally accepted value of between
-0.98 and 1.03, as defined in the `IAAO Standard on Ratio Studies`_
-Section 9.2.7. Higher PRD values indicate regressivity in assessment.
-.. _IAAO Standard on Ratio Studies: https://www.iaao.org/media/standards/Standard_on_Ratio_Studies.pdf
-.. note::
-The IAAO recommends trimming outlier ratios before calculating PRD,
-as it is extremely sensitive to large outliers. PRD is being deprecated in
-favor of PRB, which is less sensitive to outliers and easier to interpret.
-:param assessed:
-A numeric vector of assessed values. Must be the same length as ``sale_price``.
-:param sale_price:
-A numeric vector of sale prices. Must be the same length
-as ``assessed``.
-:type assessed: numeric
-:type sale_price: numeric
-:return: A numeric vector containing the PRD of the input vectors.
-:rtype: float
-:Example:
-.. code-block:: python
-# Calculate PRD:
-import assesspy as ap
-ap.prd(ap.ratios_sample().assessed, ap.ratios_sample().sale_price)
-"""
-assessed = np.array(assessed)
-sale_price = np.array(sale_price)
-# Input checking and error handling
-check_inputs(assessed, sale_price)
-ratio = assessed / sale_price
-prd = ratio.mean() / np.average(a=ratio, weights=sale_price)
-return prd
-def prb(assessed, sale_price, round=None):
-r"""
-PRB is an index of vertical equity that quantifies the
-relationship betweem ratios and assessed values as a percentage. In
-concrete terms, a PRB of 0.02 indicates that, on average, ratios increase
-by 2\% whenever assessed values increase by 100 percent.
-PRB is centered around 0 and has a generally accepted value of between
--0.05 and 0.05, as defined in the `IAAO Standard on Ratio Studies`_
-Section 9.2.7. Higher PRB values indicate progressivity in assessment,
-while negative values indicate regressivity.
-.. _IAAO Standard on Ratio Studies: https://www.iaao.org/media/standards/Standard_on_Ratio_Studies.pdf
-.. note: PRB is significantly less sensitive to outliers than PRD or COD.
-:param assessed:
-A numeric vector of assessed values. Must be the same
-length as ``sale_price``.
-:param sale_price:
-A numeric vector of sale prices. Must be the same length
-as ``assessed``.
-:param round:
-Indicate desired rounding for output.
-:type assessed: numeric
-:type sale_price: numeric
-:type round: int
-:return: A numeric vector containing the PRB of the input vectors.
-:rtype: float
-:Example:
-.. code-block:: python
-# Calculate PRB:
-import assesspy as ap
-ap.prb(ap.ratios_sample().assessed, ap.ratios_sample().sale_price)
-"""
-assessed = np.array(assessed)
-sale_price = np.array(sale_price)
-# Input checking and error handling
-check_inputs(assessed, sale_price)
-ratio = assessed / sale_price
-median_ratio = np.median(ratio)
-lhs = (ratio - median_ratio) / median_ratio
-rhs = np.log(((assessed / median_ratio) + sale_price) / 2) / np.log(2)
-lhs = np.array(lhs)
-rhs = np.array(rhs)
-prb_model = sm.OLS(lhs, rhs).fit()
-prb_val = float(prb_model.params)
-prb_ci = prb_model.conf_int(alpha=0.05)[0].tolist()
-if round is not None:
-out = {"prb": np.round(prb_val, round), "95% ci": np.round(prb_ci, round)}
-else:
-out = {"prb": prb_val, "95% ci": prb_ci}
-return out
-# Functions to determine whether assessment fairness criteria has been met
-def cod_met(x):
-return 5 <= x <= 15
-def prd_met(x):
-return 0.98 <= x <= 1.03
-def prb_met(x):
-return -0.05 <= x <= 0.05
-import assesspy as ap
-ap.prb(ap.ratios_sample().assessed, ap.ratios_sample().sale_price)
-# Import necessary libraries
-import pandas as pd
-import pkg_resources
-# Load pre-made ratios sample data.
-def ratios_sample():
-"""
-This sample was take from Evanston and New Trier in 2019. Ratios are
-calculated using assessor certified (post-appeal) fair market values.
-:return:
-A data frame with 979 observation and 4 variables:
-======================== =======================================================
-**assessed** (`float`)   The fair market assessed value predicted by CCAO
-assessment models, including any successful appeals
-**sale_price** (`float`) The recorded sale price of this property
-**ratio** (`float`)      Sales ratio representing fair market value / sale price
-**town** (`object`)      Township name the property is in
-======================== =======================================================
-:rtype: DataFrame
-"""
-stream = pkg_resources.resource_stream(__name__, "data/ratios_sample.parquet")
-return pd.read_parquet(stream)
-# Import necessary libraries
-import numpy as np
-import pandas as pd
-from pandas.api.types import is_numeric_dtype
-def check_inputs(*args):
-out = [""]
-for x in args:
-# *args passed into *args can created nested tuples - unnest
-if isinstance(x, tuple):
-args = x
-for x in args:
-if type(x) == pd.core.frame.DataFrame:
-raise Exception("Input cannot be a dataframe.")
-check = pd.Series(x)
-if not is_numeric_dtype(check):
-raise Exception("All input vectors must be numeric.")
-if check.isnull().values.any():
-out.append("\nInput vectors contain null values.")
-if len(check) <= 1:
-out.append("\nAll input vectors must have length greater than 1.")
-if not all(np.isfinite(check) | check.isnull()):
-out.append("\nInfinite values in input vectors.")
-if any(check == 0):
-out.append("\nInput vectors cannot contain values of 0.")
-out = set(out)
-if len(out) > 1:
-raise Exception("".join(map(str, out)))
-# Import necessary libraries
-import numpy as np
-import statsmodels.api as sm
-from .utils import check_inputs
-# COD, PRD, PRB functions
-def cod(ratio):
-"""
-COD is the average absolute percent deviation from the
-median ratio. It is a measure of horizontal equity in assessment.
-Horizontal equity means properties with a similar fair market value
-should be similarly assessed.
-Lower COD indicates higher uniformity/horizontal equity in assessment.
-The IAAO sets uniformity standards that define generally accepted ranges
-for COD depending on property class. See `IAAO Standard on Ratio Studies`_
-Section 9.1, Table 1.3 for a full list of standard COD ranges.
-.. _IAAO Standard on Ratio Studies: https://www.iaao.org/media/standards/Standard_on_Ratio_Studies.pdf
-.. note::
-The IAAO recommends trimming outlier ratios before calculating COD,
-as it is extremely sensitive to large outliers. The typical method used is
-dropping values beyond 3 * IQR (inner-quartile range). See
-`IAAO Standard on Ratio Studies`_ Appendix B.1.
-:param ratio:
-A numeric vector of ratios centered around 1, where the
-numerator of the ratio is the estimated fair market value and the
-denominator is the actual sale price.
-:type ratio: numeric
-:return: A numeric vector containing the COD of ``ratios``.
-:rtype: float
-:Example:
-.. code-block:: python
-# Calculate COD:
-import assesspy as ap
-ap.cod(ap.ratios_sample().ratio)
-"""
-# Input checking and error handling
-check_inputs(ratio)
-ratio = np.array(ratio)
-n = ratio.size
-median_ratio = np.median(ratio)
-cod = 100 / median_ratio * (sum(abs(ratio - median_ratio)) / n)
-return cod
-def prd(assessed, sale_price):
-"""
-PRD is the mean ratio divided by the mean ratio weighted by sale
-price. It is a measure of vertical equity in assessment. Vertical equity
-means that properties at different levels of the income distribution
-should be similarly assessed.
-PRD centers slightly above 1 and has a generally accepted value of between
-0.98 and 1.03, as defined in the `IAAO Standard on Ratio Studies`_
-Section 9.2.7. Higher PRD values indicate regressivity in assessment.
-.. _IAAO Standard on Ratio Studies: https://www.iaao.org/media/standards/Standard_on_Ratio_Studies.pdf
-.. note::
-The IAAO recommends trimming outlier ratios before calculating PRD,
-as it is extremely sensitive to large outliers. PRD is being deprecated in
-favor of PRB, which is less sensitive to outliers and easier to interpret.
-:param assessed:
-A numeric vector of assessed values. Must be the same length as ``sale_price``.
-:param sale_price:
-A numeric vector of sale prices. Must be the same length
-as ``assessed``.
-:type assessed: numeric
-:type sale_price: numeric
-:return: A numeric vector containing the PRD of the input vectors.
-:rtype: float
-:Example:
-.. code-block:: python
-# Calculate PRD:
-import assesspy as ap
-ap.prd(ap.ratios_sample().assessed, ap.ratios_sample().sale_price)
-"""
-assessed = np.array(assessed)
-sale_price = np.array(sale_price)
-# Input checking and error handling
-check_inputs(assessed, sale_price)
-ratio = assessed / sale_price
-prd = ratio.mean() / np.average(a=ratio, weights=sale_price)
-return prd
-def prb(assessed, sale_price, round=None):
-r"""
-PRB is an index of vertical equity that quantifies the
-relationship betweem ratios and assessed values as a percentage. In
-concrete terms, a PRB of 0.02 indicates that, on average, ratios increase
-by 2\% whenever assessed values increase by 100 percent.
-PRB is centered around 0 and has a generally accepted value of between
--0.05 and 0.05, as defined in the `IAAO Standard on Ratio Studies`_
-Section 9.2.7. Higher PRB values indicate progressivity in assessment,
-while negative values indicate regressivity.
-.. _IAAO Standard on Ratio Studies: https://www.iaao.org/media/standards/Standard_on_Ratio_Studies.pdf
-.. note: PRB is significantly less sensitive to outliers than PRD or COD.
-:param assessed:
-A numeric vector of assessed values. Must be the same
-length as ``sale_price``.
-:param sale_price:
-A numeric vector of sale prices. Must be the same length
-as ``assessed``.
-:param round:
-Indicate desired rounding for output.
-:type assessed: numeric
-:type sale_price: numeric
-:type round: int
-:return: A numeric vector containing the PRB of the input vectors.
-:rtype: float
-:Example:
-.. code-block:: python
-# Calculate PRB:
-import assesspy as ap
-ap.prb(ap.ratios_sample().assessed, ap.ratios_sample().sale_price)
-"""
-assessed = np.array(assessed)
-sale_price = np.array(sale_price)
-# Input checking and error handling
-check_inputs(assessed, sale_price)
-ratio = assessed / sale_price
-median_ratio = np.median(ratio)
-lhs = (ratio - median_ratio) / median_ratio
-rhs = np.log(((assessed / median_ratio) + sale_price) / 2) / np.log(2)
-lhs = np.array(lhs)
-rhs = np.array(rhs)
-prb_model = sm.OLS(lhs, rhs).fit()
-prb_val = float(prb_model.params)
-prb_ci = prb_model.conf_int(alpha=0.05)[0].tolist()
-if round is not None:
-out = {"prb": np.round(prb_val, round), "95% ci": np.round(prb_ci, round)}
-else:
-out = {"prb": prb_val, "95% ci": prb_ci}
-return out
-# Functions to determine whether assessment fairness criteria has been met
-def cod_met(x):
-return 5 <= x <= 15
-def prd_met(x):
-return 0.98 <= x <= 1.03
-def prb_met(x):
-return -0.05 <= x <= 0.05
-ap.prb(ap.ratios_sample().assessed, ap.ratios_sample().sale_price)
-# read the contents of your README file, remove image
-from pathlib import Path
-from setuptools import find_packages, setup
-this_directory = Path(__file__).parent
-image = '<a href="https://github.com/ccao-data/assesspy/tree/main"><img src="docs/images/logo.png" align="right" height="139"/></a>'
-long_description = (this_directory / "README.md").read_text().replace(image, '')
-setup(
-name="assesspy",
-version="1.0.2",
-description="General purpose Python package for measuring assessment performance",
-long_description=long_description,
-long_description_content_type='text/markdown',
-url="https://github.com/ccao-data/assesspy",
-author="CCAO",
-author_email="assessor.data@cookcountyil.gov",
-license="AGPL-3",
-packages=find_packages(),
-install_requires=[
-"pandas",
-"pyarrow",
-"numpy",
-"scipy",
-"statsmodels"
-],
-setup_requires=["pytest-runner"],
-tests_require=["pytest", "pytest-cov"],
-include_package_data = True,
-package_data={
-"": ["*.parquet"],
-}
-)
-pip install assesspy
-install.package("assesspy")
diff --git a/.gitignore b/.gitignore
index e4b39e1..d29ffb7 100644
--- a/.gitignore
+++ b/.gitignore
@@ -22,3 +22,4 @@ __pycache__/
 venv/
 .Rproj.user
 .tox
+.Rhistory