0

I have a huge data-set of a thesis experiment that needed 15 days to collect. The Experiment is repeated 15 times. I wanted to run friedman_test to check whether my data was statistically different or not. I used the following code snipet 

Data <- read.csv(file = 'myR/statistics.csv') 
Data %>% group_by(algorithm) %>% get_summary_stats(MutationScore, type = "common")
res.fried <- Data %>% friedman_test(MutationScore ~ algorithm |CUT)

My data is of the following shape, Sample data As you can see,the experiment was repeated. The green is the first run and the yellow is the second run.

When I try to run the above R code, I got the following error Exception: not an unreplicated complete block design

I read on stack overflow that my data could be invalid, or I need to use the matrix. However, I am very new to R and was stuck in this for a week. I could not help it out.

So, how can I ignore the invalid rows ? and How I can fix my friedman_test issue ?

I appreciate your help.

Thanks Hadi

Hadi Awad
  • 23
  • 4
  • Could you post a reproducible example? Otherwise, try calling the Friedman test without the pipe as passing it as the `data` argument because. For the reproducible example, you can `dput(Data)` and post the output here – csgroen Apr 26 '20 at 01:14
  • Thank @csgroen . I think the issue was that Friedman test is run on the mean (or median) data for multiple subjects (blocks). So , I need to calculate the median for my blocks ( since I have a repeated experiment) and I will try again. – Hadi Awad Apr 26 '20 at 19:36

0 Answers0