Open zhujun98 opened 5 years ago
I think Python can do the conversion easily enough, e.g. https://docs.python.org/3/library/datetime.html#datetime.datetime.utcfromtimestamp
The bit we really need is a reasonable timestamp in the files. Taking the most recent control data timestamp is not a robust way to get a meaningful timestamp. We've asked ITDM to record information from the time server in the data files, but I don't know where that's got to.
Yes, we can easily do that in Python. I simply copied the email content that I sent to SCS people since they are trying to get the timestamp from HDFView, so a web tool would be an easy way for copy-and-paste.
You can do as you say @zhujun98 as a workaround, but we are depending on an update of the file format to have this feature reliably usable. I contacted IT again on the matter, wait and see...
It would be useful to get the UTC time for each train ID. Sometimes the scientist would like to know exactly when the data is taken.
As far as I understood, a timestamp is only attached to slow data and each slow data has a different timestamp. Also, some slow data may only be updated several days before this experiment and we need to figure out which timestamp is the correct one.
A possible solution in karabo_data would be scan all the slow data for a train and sort the timestamp and convert the latest one to UTC time. Now we can only do it by hand, for example, at https://www.epochconverter.com/.