May I know how to interpret these two copulas refering to my data? Hope someone can explain to me how to interpret my results?
Thank you in advance.
It would be useful to plot your data first to see qualitatively the kind of dependencies.
Therefore, I advise to transform your original data into the respective empirical quantiles. If you have N pairs of data, the lowest value of the variable "solar radiation" gets the value 1/(N+1), the second smallest (2/(N+1),...and so on. You can do the same for the variable "sunshine hours" and then use these data pairs to generate a scatter plot.
You mentioned that for one station you got the Frank-Copula and for the other one the Clayton-Copula. Does it mean these Copulas provided the best fit according to the data?
Or that other copulas were (according to a certain significance level) explicitly rejected? Which Copulas were actually tested?
You may look the results section of the paper in the bellow url:
https://www.researchgate.net/publication/263272526_Simultaneous_Stochastic_Simulation_of_Monthly_Mean_Daily_Global_Solar_Radiation_and_Sunshine_Duration_Hours_Using_Copulas
Article Simultaneous Stochastic Simulation of Monthly Mean Daily Glo...
Thank you Dr Thomas for your answer and Dr Javad for your journal.
Dr. referring to your statement about dependency between variables, it’s possible if I just calculate the Pearson coefficient or Kendell’s tau value? And do we need to have higher dependency between chosen variable?
- Badges
- Science topic
- Similar topics
- References