I have determined that the most applicable post-hoc statistical analysis following a one-way ANOVA of my data is Dunnett's test. I have performed this using R in the past, however I am now restricted to python due to the packages I have used to automate my workflow (automatic analysis of large data quantities).
I have found several packages (eg. sci-kit, tukeyHSD) that supply multiple different post-hoc tests however none of them include Dunnett's. For example I can easily perform a one-way ANOVA in scipy:
import scipy.stats as stats
# Made up data
a = [10, 12, 10, 14, 18] # Control
b = [15, 14, 18, 10, 38]
c = [20, 22, 23, 34, 20]
d = [50, 48, 42, 51, 51]
stats.f_oneway(a, b, c, d)
> F_onewayResult(statistic=26.92639734366354, pvalue=1.7207487532445122e-06)
However I want a multiple comparison analysis following this comparing to a single control group with normally distributed data (n~1000). I am aware of Rpy2, however I wish to perform this on a machine without using a docker. Any suggestions on capable packages?
(I am also a biologist with pretty basic scripting knowledge so there's a large possibility that I'm missing something basic here)