Error in Data Frame

[mohanwugupta]

Hi,

I'm trying to run the powerstate_preprocessing function. I ran it and error occurred which I traced to the hours function. The exact error I'm getting is: "Error in data.frame(..., check.names = FALSE) :
arguments imply differing number of rows: 11, 1963". It is occurring because on line 10 inside the map function is transforming the hours variable into 11 observations versus the needed 1963 observations. This is a problem on line 14 when cbind is applied and that's when the error occurs.

Thoughts?

[mariasimoneau]

Thanks for your post and for pointing this out. We continue to work on the code base and this may get fixed as part of an effort to re-write code but there's not a particular schedule for that. In the mean time, we welcome your contributions. Thanks again.

[mohanwugupta]

Ah okay, I understand. I'll try and come with a solution in the coming weeks.

[mtpan]

Hi,

I've come up with a solution for the hours.R function only in the context of powerstate_preprocessing function. I have not tested how it works in other functions yet. Basically, I replaced the mapper function with a for loop to get it to work. I'm not sure why this solution works but it seems to work across multiple datasets.
hours.txt

[brainfood]

So this comes long after the last comment, but here's a modified version of the hours() function that works for me across several of the pre-processing functions.

hours = function(timestamps){ 
    days = timestamps %>% unlist %>% as.POSIXlt(origin = "1970-01-01") %>%
        as.character %>%
        map(function(x){if(is.na(x)){return(c("NA","NA"))}else{return(strsplit(x, " "))}}) %>%
        unlist %>%
        matrix(nrow=2) %>%
        t
    hours =
        days[,2] %>% 
        map(strsplit, split=":") %>% 
        unlist() %>% 
        as.numeric() %>% #Convert from factor
        matrix(ncol=3, byrow = TRUE) %>% 
        data.frame() %>% 
        dplyr::rename(hour=X1, min=X2, sec=X3) %>%
        apply(1, function(times) sum(times * c(1, 1/60, 1/3600)))
    days = days[,1]
    output = as_data_frame(cbind(hours=hours, days=days))
    output["hours"] = lapply(output["hours"], as.numeric)
    return(output)
}