Last chance! 50% off unlimited learning
Sale ends in
vectorplot displays vector fields from Raster
objects using arrows.
streamplot displays streamlines with a procedure inspired by
the FROLIC algorithm (see references): for each point (droplet)
of a jittered regular grid, a short streamline portion
(streamlet) is calculated by integrating the underlying vector
field at that point. The main color of each streamlet indicates local
vector magnitude (slope). Streamlets are composed of points
whose sizes, positions and color degradation encode the local vector
direction (aspect).
# S4 method for Raster
vectorplot(object, layers,
narrows=2e3, lwd.arrows=0.6, col.arrows='black',
length=unit(1e-2, 'npc'),
maxpixels=1e5, region=TRUE, margin=FALSE,
isField=FALSE, reverse=FALSE,
unit='radians', scaleSlope=TRUE,
aspX=0.08, aspY=aspX,
key.arrow = NULL,
...)# S4 method for RasterStack
vectorplot(object, layers,
narrows=2e3, lwd.arrows=0.6, col.arrows='black',
length=unit(1e-2, 'npc'),
maxpixels=1e5, region=TRUE, margin=FALSE,
isField=FALSE, reverse=FALSE,
unit='radians', scaleSlope=TRUE,
aspX=0.08, aspY=aspX,
key.arrow = NULL,
uLayers, vLayers, ...)
# S4 method for Raster
streamplot(object, layers,
droplet = list(), streamlet = list(),
par.settings = streamTheme(),
colorkey = FALSE,
isField = FALSE, reverse = FALSE,
parallel = TRUE, mc.cores = detectCores(), cl = NULL,
...)
# S4 method for RasterStack
streamplot(object, layers,
droplet = list(), streamlet = list(),
par.settings=streamTheme(),
colorkey = FALSE,
isField = FALSE, reverse = FALSE,
parallel = TRUE, mc.cores = detectCores(), cl = NULL,
...)
A Raster object. If isField=FALSE the vector
field is calculated internally from the result of
terrain.
A numeric or character which should indicate the layers to be displayed.
A numeric, number of cells to be shown if
region=TRUE or if region is a Raster* object.
A numeric, number of arrows.
Numeric, width of the lines of the arrows
character, color of the arrows
Unit, extent of the arrow head.
Logical, if TRUE two marginal graphics show the summaries of the object.
Logical or numeric. If TRUE the slope (vector
magnitude) is scaled (but not centered) with its standard
deviation. If it is a numeric, the slope is scaled with this value.
It is not used if isField='dXY'.
Numeric. Multipliers to convert the slope values into
horizontal (aspX) and vertical (aspY) displacements.
A reference (or legend) vector. If is not NULL, it is a list with two named
components, size, a numeric to define the length of the arrow
(default 1), and label, a character to define the label
(default '').
Numeric, indexes of layers with horizontal and
vertical components, respectively, when isField='dXY' and the
RasterStack has more than 2 layers. If missing, the horizontal
components are the layers of the first half of the object, and the
vertical components are the layers of the second half.
A list whose elements define the droplet configuration:
cropExtent: Percentage of the object extent to be cropped (default .97) to avoid droplets at boundaries
pc: A numeric. It is the percentage of cells used to compute
droplets. Its default value is 0.5. Therefore, only the 0.5% of the
cells are used. For example, if you use a Raster with 180 rows and 360
columns (64800 cells), with this default value streamplot will
produce 324 droplets.
A list whose elements define the streamlet configuration:
L: length of the streamlet (number of points, default 10)
h: streamlet calculation step (default
mean(res(object))).
A list to define the graphical parameters. For
streamplot there is an specific theme, streamTheme.
A logical specifying whether a color key is to be drawn
alongside the plot (default is FALSE), or a list describing
the color key (see the colorkey section of the help page of
levelplot for details).
Logical, TRUE (default) to use parallel
package.
a cluster object. Read the help page of
parLapply for details.
The number of cores to use if parallel=TRUE and
no cl is provided. Read the help page of mclapply
for details.
Logical, if TRUE the region is displayed as the
background using levelplot. It can be a Raster* with
the same extent and resolution as object.
If TRUE the object is a vector field. Thus,
object must be a Raster* with two layers, slope (local vector
magnitude) and aspect (local vector direction), in this order. The
aspect angle, following the conventions of
raster::terrain, must be computed clockwise from 0 (due
north). The slope layer will be used as the background if region
is TRUE.
If isField='dXY' object must be a Raster* with two
layers representing the horizontal and the vertical components,
respectively. The slope is computed and used as the background if
region is TRUE.
Character, angle units of the aspect layer if
isField=TRUE: 'radians' or 'degrees'.
Logical, if isField=TRUE and reverse=TRUE,
arrows or streamlets go against the direction of the gradient.
Additional arguments for levelplot
Oscar Perpiñán Lamigueiro
R. Wegenkittl and E. Gröller, Fast Oriented Line Integral Convolution for Vector Field Visualization via the Internet, Proceedings IEEE Visualization ’97, 1997, http://www.cg.tuwien.ac.at/research/vis-dyn-syst/frolic/frolic_crc.pdf
panel.arrows,
levelplot,
terrain,
mclapply,
parLapply
if (FALSE) {
library(raster)
proj <- CRS('+proj=longlat +datum=WGS84')
df <- expand.grid(x=seq(-2, 2, .01), y=seq(-2, 2, .01))
df$z <- with(df, (3*x^2 + y)*exp(-x^2-y^2))
r1 <- rasterFromXYZ(df, crs=proj)
df$z <- with(df, x*exp(-x^2-y^2))
r2 <- rasterFromXYZ(df, crs=proj)
df$z <- with(df, y*exp(-x^2-y^2))
r3 <- rasterFromXYZ(df, crs=proj)
s <- stack(r1, r2, r3)
names(s) <- c('R1', 'R2', 'R3')
vectorplot(r1)
vectorplot(r2, par.settings=RdBuTheme())
vectorplot(r3, par.settings=PuOrTheme())
## scaleSlope, aspX and aspY
vectorplot(r1, scaleSlope=FALSE)
vectorplot(r1, scaleSlope=1e-5)
vectorplot(r1, scaleSlope=5e-6, alpha=0.6)
vectorplot(r1, scaleSlope=TRUE, aspX=0.1, alpha=0.6)
vectorplot(r1, scaleSlope=TRUE, aspX=0.3, alpha=0.3)
## Reference vector
# Default size (1)
vectorplot(r1, region = FALSE,
key.arrow = list(label = 'm/s'))
vectorplot(r1, region = FALSE,
key.arrow = list(size = 2, label = 'm/s'))
## A vector field defined with horizontal and vertical components
u <- v <- raster(xmn=0, xmx=2, ymn=0, ymx=2, ncol=1e3, nrow=1e3)
x <- raster::init(u, fun='x')
y <- raster::init(u, fun='y')
u <- y * cos(x)
v <- y * sin(x)
field <- stack(u, v)
names(field) <- c('u', 'v')
vectorplot(field, isField='dXY', narrows=5e2)
## We can display both components as the background
vectorplot(field, isField='dXY', narrows=5e2, region=field)
## It is also possible to use a RasterStack
## with more than 2 layers when isField='dXY'
u1 <- cos(y) * cos(x)
v1 <- cos(y) * sin(x)
u2 <- sin(y) * sin(x)
v2 <- sin(y) * cos(x)
field <- stack(u, u1, u2, v, v1, v2)
names(field) <- c('u', 'u1', 'u2', 'v', 'v1', 'v2')
vectorplot(field, isField='dXY',
narrows=300, lwd.arrows=.4,
par.settings=BTCTheme(),
layout=c(3, 1))
## uLayer and vLayer define which layers contain
## horizontal and vertical components, respectively
vectorplot(field, isField='dXY',
narrows=300,
uLayer=1:3,
vLayer=6:4)
##################################################################
## Streamplot
##################################################################
## If no cluster is provided, streamplot uses parallel::mclapply except
## with Windows. Therefore, next code could spend a long time under
## Windows.
streamplot(r1)
## With a cluster
hosts <- rep('localhost', 4)
cl <- parallel::makeCluster(hosts)
palRed <- RColorBrewer::brewer.pal(n = 5, name = 'Reds')
streamplot(r2, cl=cl,
par.settings=streamTheme(symbol= palRed))
parallel::stopCluster(cl)
## Without parallel
palGreen <- RColorBrewer::brewer.pal(n = 5, name = 'Greens')
streamplot(r3, parallel=FALSE,
par.settings=streamTheme(symbol = palGreen))
## Configuration of droplets and streamlets
streamplot(s, layout=c(1, 3), droplet=list(pc=.2), streamlet=list(L=20),
par.settings=streamTheme(cex=.6))
}
Run the code above in your browser using DataLab