search

AndyFlury13 / lid

Scripts written during my work for Professor Ani Jayakaran and the Washington Stormwater Center. Analysis of soil moisture data in a raingarden study.
0 stars 0 forks source link

P-Values at Zero for ANOVA #1

Open bleonard314 opened 3 years ago

bleonard314 commented 3 years ago

I think I found the issue with those p-values. It looks like when you melt to overwrite “step_data” on line 328 you do so with the whole data frame which makes “Date” a treatment variable. Since values for date are converted to numeric values that are orders of magnitude different than the treatments, this drives the p-values down to basically 0. If you replace with step_data[-1] this brings the p-values up slightly.

AndyFlury13 commented 3 years ago

Hey Ben,

I took a look at the dataframe step_data and was able to make some changes. I did remove date from the melted dataframe and I got a much lower F-statistic, though my P value is still far below 0. I may try lm() instead...