

In [2]:

    
import numpy as np
from sklearn.preprocessing import StandardScaler
from sklearn.linear_model import LinearRegression
from sklearn.pipeline import Pipeline
import pandas as pd

%matplotlib inline
import matplotlib.pyplot as plt
plt.style.use('ggplot')

from basis_expansions import (Binner, Polynomial, 
                              LinearSpline, CubicSpline,
                              NaturalCubicSpline)









    



/Users/matthewdrury/anaconda/envs/py36/lib/python3.6/site-packages/matplotlib/__init__.py:913: UserWarning: axes.color_cycle is deprecated and replaced with axes.prop_cycle; please use the latter.
  warnings.warn(self.msg_depr % (key, alt_key))

Basis Expansions Vignette

Introduction

Installation

To install, clone the repository and run setup.py:

git clone https://github.com/madrury/basis-expansions
cd basis-expansions
pip install .

Quick Start

Basis Expansions



In [11]:

    
def plot_basis(axs, basis_expansion, t):
    basis = basis_expansion.fit_transform(t)
    for idx, ax in enumerate(axs.flatten()):
        ax.plot(t, basis[:, idx])

Binning Expansion



In [18]:

    
binner = Binner(min=0, max=1, n_cuts=9)

fig, ax = plt.subplots(5, 2, figsize=(12, 6))
plot_basis(ax, binner, np.linspace(0, 1, num=1000))
fig.tight_layout()

Polynomial Expansions



In [22]:

    
poly = Polynomial(degree=10)

fig, ax = plt.subplots(2, 5, figsize=(12, 4))
plot_basis(ax, poly, np.linspace(-2, 2, num=1000))
fig.tight_layout()

Piecewise Linear Expansions



In [27]:

    
lin = LinearSpline(min=0, max=1, n_knots=9)

fig, ax = plt.subplots(2, 5, figsize=(12, 4))
plot_basis(ax, lin, np.linspace(0, 1, num=1000))
fig.tight_layout()

Natural Cubic Spline Expansions



In [30]:

    
cub = NaturalCubicSpline(min=0, max=1, n_knots=11)

fig, ax = plt.subplots(2, 5, figsize=(12, 4))
plot_basis(ax, cub, np.linspace(0, 1, num=1000))
fig.tight_layout()



In [ ]: