Hey Akbar,

I am not sure if I got all your questions as there were quite a lot. So I am just trying to answer what I consider your main questions.

First of all, the data you provided does not include a column named simulated so I am not able to fully reproduce your example.

If I understood you correct you are aiming at calculating the RMSE and different other statistics for different groups in your data (e.g., year, month, day, ...). Your data contains a time column called DAYTIME which contains all needed information. I would do the following:

# load necessary packages

library(tidyverse)

# Split your DAYTIME column into three single columns containg Year, Month, Day

df % group_by(Year) %>% select(Year,OBSERVED) %>% summarize(SQRT = sqrt(OBSERVED), Mean = mean(OBSERVED))

In this manner you should be able to calculate all statistics you are looking for. Simply add your SIMULATED variable (which is missing in your data file) to the select function:

df %>% group_by(Year) %>% select(Year,OBSERVED,SIMULATED) %>% summarize(RMSE = sqrt(mean(SIMULATED - OBSERVED)^2), Mean = mean(SIMULATED-OBSERVED))

Hope that helps.

All the best,

Fabian

thank you dear Fabian

I have another question too, with multiple observed and also multiple simulated columns what should I do? how I can fix this issue too I mean to calculate statistic between observed columns and simulated ones( all four other columns)? Fabian Greb

Can you specify your question? Do you want to merge the several simulated or observed columns before calculating any statistics or is it that you want to calculate statistics like:

simulated1 - observed1

simulated2 - observed2

etc...

actually, I've merged them before, I just want to calculate different statistics between them see the below-attached file and you will find what I'm saying 

Hi,

I am sorry but I still do not really get your problem. I see 4 different observed variables obs_17 - obs_63. Please try to precisley describe want you want to do with them.