By running 10,000 permutations (shuffles) of A and B, generate a null distribution of differences between A and B.
computer science
Description
Q2 By running
10,000 permutations (shuffles) of A and B, generate a null distribution of
differences between A and B.
If you find that this takes some
time, you might want to have your for loop print out its current iteration so
you can monitor its progress
Q3 Calculate the p-value of your observed difference. To do this, simply count the number of dfference values in
the null distribution that are more extreme than the observed difference (in
EITHER direction, positive or negative), and express this as a proportion of
10,000!
Q4 Plot this
null distribution along with a vertical line where the observed difference
was.
Include plot labels and a legend
Q5 Define a
function to carry out this whole process
Your
function should take as its input arguments:
- The number of
permutations to do
- The population
mean of A and B
- The population
standard deviation of A and B
- The sample size
of A and B
It should
produce both plots generated above, with all input parameters described in the
title
Furthermore,
it should output:
- The observed
difference of your samples
- The null
distribution of differences from your permutations
Set the
above values as the default values for your input parameters (values the input
variables take if you do not specify them in the input argument list) by using
this function definition:
def run_permtest(nperm=10000 , u=[100,102] , s=[4,4] , n=[60,60]):
Q6 Using your
function above, see what happens when you eliminate the effect (i.e. population
means of A and B are equal), but using the default values for the other
parameters
Assign your outputs to unique
variables so that you can use them again later.
