Hi all,
I have a data file containing earthquake records of the following format: Year MoDaHrMnSecs Lat Lon Dep M 2015 0416180744.30 35.1460 26.8880 0.00 6.1 MS 2015 0416181417.60 35.1280 26.6650 1.50 3.6 2015 0416182112.50 35.0360 26.6920 0.20 3.8 2015 0416182354.70 35.1120 26.7660 0.00 3.5 2015 0416183004.50 35.0190 26.6770 9.60 3.4 [...]
In order to search the record with the maximum magnitude, I use the command:
gmt info -Eh5 -h1 datafile
but the results differ in GMT 6.0.0 (FC 31, fedora yum repo) and 5.4.4 (CentOS 7, epel repo):
in 6.0.0: 2015 416180744.3 35.146 26.888 0 6.1 MS
in 5.4.4: gmtinfo: Mismatch between actual (6) and expected (7) fields near line 3
If I bypass the 7th column with: sed '2s/ MS\ *$//' datafile | gmt info -Eh5 -h1
I still receive a different output in GMT 5: 2015 0416180744.30 35.1460 26.8880 0.00 6.1
Is there a way to ask from gmtinfo not to ignore the leading zero of second column (when it exists), since I need it later to calculate time differences and plot accordingly?
Thanks in advance
There are some problems here that are difficult to overcome: First, your time data is not in any ISO-recognizable format [https://en.wikipedia.org/wiki/ISO_8601] and hence it is (or cannot be) interpreted as a valid time-stamp. Your time is actually placed across two columns so it is even worse (a time stamp is normally something like yyyy-mm-ddThh:mm:ss.xxxx) It is thus seen as a floating point number and hence to leading 0 is lost in the conversion. You may have software that knows about this particular two-column arrangement but GMT cannot do that. The 5.4.4/6.0.0 difference is tied to a change in the GMT table record format where all text is expected to be in an optional trailing text after all numerical columns. Thus, it does not bother GMT 6 that you have an occasional trailing text (MS) in some records but not in others, but it really bothers GMT 5 who will try to read that MS and obtain 0 or NaN.
One workaround would be to awk your file and place the time information into the trailing text. To ensure it is seen as text, maybe print it with a leading T or some other text string, then after you have done your GMT processing you can reformat to whatever final format you need.