specification-curve 0.1.1

Quickstart

Running

from specification_curve import specification_curve as sc
from specification_curve import example as scdata
df = scdata.load_example_data1()
y_endog = 'y2'
x_exog = ['x1', 'x2']
controls = ['c1', 'c2', 'group1', 'group2']
df_r = sc.spec_curve(df, y_endog, x_exog, controls,
                     cat_expand=['group2'])

produces

Grey squares (black lines when there are many specifications) show whether a variable is included in a specification or not. Blue markers and error bars show whether the coefficient is significant (0.05).

Here’s another example:

from specification_curve import specification_curve as sc
import numpy as np
n_samples = 300
np.random.seed(1332)
x_1 = np.random.random(size=n_samples)
x_2 = np.random.random(size=n_samples)
x_3 = np.random.random(size=n_samples)
x_4 = np.random.randint(2, size=n_samples)
y = (0.8*x_1 + 0.1*x_2 + 0.5*x_3 + x_4*0.6 +
     + 2*np.random.randn(n_samples))
df = pd.DataFrame([x_1, x_2, x_3, x_4, y],
                  ['x_1', 'x_2', 'x_3', 'x_4', 'y']).T
# Set x_4 as a categorical variable
df['x_4'] = df['x_4'].astype('category')
df_r = sc.spec_curve(df, y_endog, x_exog, controls,
                     cat_expand=['x_4'])

Features

These examples use the first set of example data:

df = edata.load_example_data1()

• Expand fixed effects into mutually exclusive groups using cat_expand

y_endog = 'y1'
x_exog = 'x1'
controls = ['c1', 'c2', 'group1', 'group2']
df_r = sc.spec_curve(df, y_endog, x_exog, controls,
                         cat_expand=['group1', 'group2'])

• Mutually exclude two variables using exclu_grp

y_endog = 'y1'
x_exog = 'x1'
controls = ['c1', 'c2', 'group1', 'group2']
df_r = sc.spec_curve(df, y_endog, x_exog, controls,
                 exclu_grps=[['c1', 'c2']])

• Use multiple independent or dependent variables

x_exog = ['x1', 'x2']
y_endog = 'y1'
controls = ['c1', 'c2', 'group1', 'group2']
df_r = sc.spec_curve(df, y_endog, x_exog, controls)

Similar Packages

In RStats, there is specr (which inspired many design choices in this package) and spec_chart. Some of the example data in this package is the same as in specr.

Credits

This package was created with Cookiecutter and the audreyr/cookiecutter-pypackage project template.

0.1.0 (2020-07-27)

• First release on PyPI.

0.1.1 (2020-08-01)

• Multiple independent, dependent, and control variables implemented as lists. Mutually exclusive control variables implemented. Expansions of categorical variables into mutually exclusive fixed effects implemented.