combine.rt: Aggregate data on square grid, pad with zeros. Faster than...
In geo: Draw and Annotate Maps, Especially Charts of the North Atlantic

Description Usage Arguments Details Value See Also

Function that smooths data so data. In each given square where the number of points exceeds certain minimum the program gives the mean, sum, variance or median of the points within the square as well as the number of points and indication if the minimum number of points was reached. For each datapoint latitude, longitude and z has to be given. The program returns the mean and mean longitude for points within a square, as well as the number of points behind that point. The program also has the possibility to remove outliers within each square. This program can be used to prepare data for the program pointkriging if the data is so dense in certain places that it disturbs the neighbourhood search or if there are +10000 datapoints.

1 2	combine.rt(lat, lon, z, grlat, grlon = 0, fun, fill = F, reg = 0, minnumber = 2, wsp = 0, wz = 0, wlat = 0, xy = F, rat = 0.2, type)

`lat`	Latitude of datapoints. ( or x coordinates)
`lon`	Longitude of datapoints. (or y coordinate)
`z`	Values at datapoints.
`grlat`	Latitude of defined grid.
`grlon`	Longitude of defined grid. If 0 grlat is a list with components \$lat and \$lon.
`fun`	Function (?).
`fill`	If fill is T the center points of squares where the number of points does not reach the minimum is set to zero. Good for data like trawlers report where it can be assumed that there is little fish where they do not try.
`reg`	List with two components \$lat and \$lon. Points outside region are not returned. The list is typically output from the program define.area.
`minnumber`	Minimum number of points needed for the square to be valid. Default value is 2.
`wsp`	Workspace. Not of interest.
`wz`	Weighing for the z values. If one is trying to find the mean tons/hour in an area is it reasonable to weigh tons/hour by the catch. If the weight is zero within a square unweighed mean is used.
`wlat`	Weighing to find the mean latitude and mean longitude. It is often reasonable to weigh the latitude and the longitude by the catch.
`xy`	Projected? Default unprojected (lat/lon) values, `FALSE`.
`rat`	Ratio for "rm.outliers". If rat is for example 0.2 20% of the points on each side are moved. i.e. the middle 60% are kept. Default value is 0.6.
`type`	Only value with description `rm.outliers`.

It can also be used to find sums or averages of points within squares.

The program works quite fast even for very large datasets.

<s-example> List with components \$lat,\$lon,\$z,\$n and \$fill. \$lat mean latitude of points within each square. \$lon mean longitude of points within each square \$z sum, mean, variance or median of data \$n Number of points behind point.

If type = "rm.outliers" \$n this component stores the number of the square corresponding to each point. \$fill If 0 there are datapoints behind the points else it is an artificial point. (fill = T) </s-example>

variogram, variofit, pointkriging, grid, geocontour.fill.