nnl

nnl function allows to extrapolate SpatialLinesDataFrame A to a SpatialLinesDataFrame B by finding nearest neighbors lines. It create points along two SpatialLinesDataFrames and it find nearest points. Then, statistics are use to select nearest lines.

Install

Install package with devtools :

devtools::install_github("naub1n/nnl")

How to use

Required arguments

Just use nnl::nnl()function.

l_A is the original SpatialLinesDataFrame

l_B is a SpatialLinesDataFrame where you want to extrapolate l_A

id_l_A an id_l_B are columns names contain unique values (IDs)

Optionals arguments

in progress

Examples

in progress

Performance

With shp_me, DCE river on Seine-Normandie.

Large SpatialLinesDataFrame (1651 elements, 10.3 Mb)

With shp_topo, BD topo V2 on Seine-Normandie

Large SpatialLinesDataFrame (295908 elements, 626.1 Mb)

Start nnl() function:

system.time(test <- nnl::nnl(l_A = shp_me, l_B = shp_topo, id_l_A = "CdMasseDEa", id_l_B = "ID",verbose = T))

Function responses :

CRS Comparaison
Projection test
Filter l_B
Create points to l_A
5163216 points created
Create points to l_B
8210254 points created
Find nearest points
8210254 points attached
Select nearest lines : Step 1
55643 lines found
Select nearest lines : Step 2
2345 lines found
Aggregate result
Select nearest l_B
Add IDs
Compare length result with l_A length
End

Result :

Large SpatialLinesDataFrame (58092 elements, 152.9 Mb)

Time result :

utilisateur     système      écoulé 
     122.21        7.05      256.55 

4 min 16 sec*

* time with i7-7500U CPU (2C4T) and 16Go RAM