Thanks Benjamin! Creating a matrix out of my two arrays works just fine.
I'll go that way for now.
Josh
On Tuesday, November 10, 2015 at 4:32:27 PM UTC-6, Benjamin Deonovic wrote:
>
> Looks like the state of Julia's "Factors" (PooledDataArrays) is still
> quite a fast evolving monster.
I found your Chi-Square test and am trying to use it. It appears to work
with one array but not with two. My steps are below:
This works:
ChisqTest([1,2,3,4])
This doesn't:
ChisqTest([1,2,3,4],[1,2,2,4])
It errors with the following:
LoadError: MethodError: `ChisqTest` has no method
Try a matrix?
Hey Joshua,
Just saw your post. I will investigate into what the issue is. I wrote this
quite a while ago when I was just learning Julia!
On Tuesday, November 10, 2015 at 9:10:10 AM UTC-6, Joshua Duncan wrote:
>
> I found your Chi-Square test and am trying to use it. It appears to work
> with
Looks like the state of Julia's "Factors" (PooledDataArrays) is still quite
a fast evolving monster. Because of this there hasn't been a proper
implementation of R's "table" function. For now, if you want to run
ChisqTest in Julia on two vectors do it on the matrix you would obtain in R
by
Okay I have figured out the issue. I will fix it so it works the way you
expected it to work. Before the fix goes live though it should work to do:
ChisqTest([1,2,3,4],[1,2,2,4], 4)
*note the 4
The issue was when I submitted the code to HypothesisTests.jl the only way
to create a contingency
My pull has been merged: https://github.com/JuliaStats/HypothesisTests.jl
On Thursday, March 5, 2015 at 8:32:32 AM UTC-6, Benjamin Deonovic wrote:
I implemented the chisquare test in julia. I made a pull request in the
HypothesisTests package. It hasn't been pulled yet, but probably will be
I implemented the chisquare test in julia. I made a pull request in the
HypothesisTests package. It hasn't been pulled yet, but probably will be
soon.
On Sunday, February 8, 2015 at 5:32:48 PM UTC-6, Arin Basu wrote:
Hi All,
Please pardon my ignorance, but how does one do chisquare test in
Hi Arin,
It would appear that there isn't a chi square test at the moment. If you
want to implement one yourself, you can compute the quantiles of a chi
squared distribution using DIstributions.jl (if you do this, please
consider submitting it to HypothesisTests.jl). Alternatively, you could
