R datetime格式问题

Question

I am currently trying to determine the time and date on the observations in my dataset. The date/timestamp is as follows: 1458024601.18659 1458024660.818 The observation are recorded ever minute. I am trying to convert the above date/time stamp into something for understandable/ interpretable. Could you please help me with this issue. Many thanks.

Answer 1

Looks like seconds, but seconds starting from when? Typically, 1970-01-01:

> x = 1458024601.18659
> as.POSIXct(x, origin="1970-01-01")
[1] "2016-03-15 06:50:01 GMT"

So if you are expecting that timestamp to be that time, we've got the origin right.

If you are expecting a date in 1946, then origin="1900-01-01" is probably what you want.

Answer 2

Since, according to your most recent post, the data is stored as a factor class, some further manipulations are required.

To convert the factor column into the required numeric class, this modification of @Spacedman's answer should work:

as.POSIXct(as.numeric(as.character(all_prices$timestamp)), origin="1970-01-01")

Answer 3

Your solution is perfect, except that i have another issue :( I tried to run this code on the data.frame that i have. Unfortunately, i keep getting this following error after running the code.

dates <- as.POSIXct(all_prices$timestamp, origin="2016-03-15") Error in as.POSIXlt.character(as.character(x), ...) : character string is not in a standard unambiguous format

Data "all_prices" is a data.frame.

class(all_prices) [1] "data.frame"

data "all_prices$timestamp" is a factor.

class(all_prices$timestamp) [1] "factor"