Geopointset is the format used by Metview to combine a set of Geopoints variables into a single entity for ease of processing. Thus, a set of observations can be grouped in the same way that fields are grouped into a fieldset variable. For a full list and details of functions and operators on geopoints, see Geopointset Functions.
Creating a geopointset
A geopointset can be created with the create_geo_set() function, which takes any number of geopoints variables as arguments, or none. Both geopoints and geopointset variables can be concatenated with a geopointset.
set1 = create_geo_set() # creates an empty set set2 = create_geo_set(g1, g2, g3) # assuming that g1,g2,g3 are geopoints variables set3 = set1 & g1 & g2 # set3 has 2 geopoints set4 = set2 & set3 # set4 has 5 geopoints
Accessing geopintset elements
The count() function returns the number of geopoints variables contained by the set.
Use the indexing operator [] to access the geopoints variables contained in a set. For example:
print(type(set4)) # geopointset print(count(set4)) # 5 g1 = set4[1] # extract the first item print(type(g1)) # geopoints print(count(g1)) # 244 (if there are 244 points in this geopoints variable)
Operations on geopointsets
As a geopointset is simply a container for geopoints variables, most operations on a geopointset are performed on each of its component geopoints. For example, the following line of code with return a new geopointset where each geopoints variable has had the cos() function applied to its values:
cgset = cos(gset)
Operations between geopointsets and numbers are performed on each geopoints, e.g.
gsetplus1 = gset + 1 # add 1 to each value in each geopoints var in gset
Operations can be performed between geopointsets and geopointsets, or geopointsets and fieldsets, as long as they both contain the same number of items, or they contain exactly one item. Otherwise, if they contain a different number of items, the computation will fail.
For example, if gset_5a and gset_5b each contain 5 geopoints variables, the following code will add each pair of geopoints variables, giving a resulting geopointset of size 5:
gsetsum_r1 = gset_5a + gset_5b # gset_5b[n] is added to gset_5a[n]
If gset_1c contains a single geopoints variable, the following code will produce a geopointset with 5 items, the result of adding gset_1c[1] to each item in gset_5a:
gsetsum_r2 = gset_5a + gset_1c # gset_1c[1] is added to each gset_5a[n]
Likewise, geopointset/fieldset operations work the same way:
gsetdiff_r1 = fc_fieldset_5 - gset_5a # gset_5a[n] is subtracted from fc_fieldset_5[n] gsetdiff_r2 = fc_fieldset_5 - gset_1c # gset_1c[1] is subtracted from each field
Filtering a geopointset
Individual geopoints variables can contain meta-data - see Geopoints for details. To select only those geopoints variables with given meta-data, use the filter() function as described in Geopointset Functions.
The Geopointset file format
The format for a geopointset file is very simply a header followed by a contatenation of geopoints files - see Geopoints for details of the format. The overall header is this:
#GEOPOINTSET
The subsequent geopoints structures should all share the same format as each other. Here's an example with 3 geopoints files inside the set:
#GEOPOINTSET #GEO # lat lon height date time value # Missing values represented by 3e+38 (not user-changeable) #DATA 69.6523 18.9057 0 20130512 0 100869.8625 63.4882 10.8795 0 20130512 0 100282.3392 63.5657 10.694 0 20130512 0 100241.1666 61.2928 5.0443 0 20130512 0 99852.18932 #GEO # lat lon height date time value # Missing values represented by 3e+38 (not user-changeable) #METADATA param=geopotential #DATA 60.82 23.5 0 20130512 600 101045.8 #GEO # lat lon height date time value # Missing values represented by 3e+38 (not user-changeable) #DATA 55.01 8.41 0 20130513 0 100949.1809 54.33 8.62 0 20130513 0 101027.9101 53.71 7.15 0 20130513 0 100846.619