Panel_data$date <- as. Panel_data <- (date = c(date6,date5,date4,date6), symbol = symbol, rskew = rskew, rkurt = rkurt, cor_skew_kurt = cor_skew_kurt ) #Recreating an example dataset with unequal dates across stocks if I understood your problem correctly this code should at least put you on the path to the right solution as it solves the issue of unequal time window length. Can you try running the below code? I have recreated an example emulating your issue. Or should I maybe use a different approach to do this? Is there a way to do this without having to define a fixed width for each date group? Mutate(cor_skew_kurt = rollapply(data = panel_data, Here is what i tried for a sample width of 20: panel_data % I tried to do it with the rollapply function from the zoo package, but since the number of different stocks is not the same for all dates, I cannot simply define width as an integer.
This means I need to compute the correlation between rskew and rkurt over all different stocks at each point in time and then calculate the time-series average afterwards. I want to calculate the time-series average of the cross-sectional correlation between the variables rskew and rkurt. Where the variable symbol depicts different stocks.
53.3, p < 0.01) were observed in the first period. 64.4 ± 22.6 years, p < 0.01) and a higher proportion of cases with SARI (85.2 vs.
0 kommentar(er)
