Small wrapper class for using R's linear regression functionality in Python, via rpy2.

Create some data:

>>> x = [1, 2, 3, 4]
>>> y = [1.2, 3, 7, 10]

Create the model. The keyword arguments you provide are used to name the variables in R, so that's how you need to refer to them in the formula:

>>> from linregress import LinearRegression
>>> m = LinearRegression(x=x, y=y, formula='y~x')

As an example of this, m2 will be the same model as m with the same data -- it's just that the keyword argument names have been changed, and so the formula is correspondingly changed:

>>> m2 = LinearRegression(dependent=x, response=y, formula='response~dependent')

Slope, intercept, R-squared, and adjusted R-squared are available as attributes:

>>> m.slope
3.0399999999999996

>>> m.intercept
-2.299999999999998

>>> m.adj_r_squared
0.97221750212404412

>>> m.r_squared
0.98147833474936275

For testing if the slope is different from some expected value (0 by default), use the slope_pval() method:

>>> m.slope_pval()
0.0093041159117684229

>>> m.slope_pval(compare_to=0)
0.0093041159117684229

>>> m.slope_pval(compare_to=3.0)
0.90465374107544172

The intercept_pval() method works similarly:

>>> m.intercept_pval()
0.10459053583417365

>>> m.intercept_pval(compare_to=0)
0.10459053583417365

>>> m.intercept_pval(compare_to=5)
0.012051074953242348

The diagnostics() method will plot the the diagnostics. Providing a filename will save the plots there; with no filename an X11 window will appear which you can close with linregress.rclose():

>>> m.diagnostics('diagnostics.png')

There is also limited support for ANCOVA (analysis of covariance) to compare slopes or intercepts of different models.

First create another model to compare:

>>> x2 = [2, 4, 7, 9]
>>> y2 = [1, 4, 6, 10]
>>> m2 = LinearRegression(x=x2, y=y2, formula='y~x')

Then run the ANCOVA. The return value is (pval of slope comparison, pval of intercept comparison):

>>> from linregress import ancova
>>> ancova(m, m2)
(0.0089121329373064283, 0.040977724663148127)