v.db.univar uses db.univar which in turn uses db.select to get the attribute values on which it calculates the statistics. This means that statistics are calculated based on the entries in the attribute table, not based on the features in the map. One attribute value is read from each line in the attribute table, whether there are no, one or several features with the category value referenced by that line, or whether any features have more than one category value. For feature-based, instead of attribute table-based, univariate statistics on attributes see v.univar. NOTES A database connection must be defined for the selected vector layer.
# show columns of attribute table connected to precipitation map v.info -c precip_30ynormals # univariate statistics on 30 years annual precipitation in NC v.db.univar precip_30ynormals column=annual Number of values: 136 Minimum: 947.42 Maximum: 2329.18 Range: 1381.76 Mean: 1289.31147058823 [...]
g.region raster=elevation -p v.random output=samples n=100 v.db.addtable samples column="heights double precision" v.what.rast samples raster=elevation column=heights v.db.select samples v.db.univar samples column=heights
and authors of r.univar.sh (Markus Neteler et al.)
Last changed: $Date$