Integrating variable plot radii in plot_metrics()

[benice2me]

Hi everyone,

For an ABA project I am currently working on I am trying to compute plot metrics using plot_metrics. The catch is that my plots have different radii (radiuses?). My plot_data, consisting of 165 rows and 16 columns, has a column containing these radii. What is the most efficient way to integrate these in the plot_metrics formula?

Thanks in advance and kind regards,
Joep

Also: first time posting on GitHub, please excuse my lack of formatting.

[Jean-Romain]

You can provide several radii (one per plot)

plot_metrics(ctg, centers, radius = center$radius)

[benice2me]

Thank you for the fast reply. Do you have a link to where this is documented? Also, is centers in your example the plot data?

[Jean-Romain]

Also, is centers in your example the plot data?

center is your source of spatial points (e.g. a shapefile) in a sp or sf format. But you can do it another way

Do you have a link to where this is documented?

Well, thechnically it is documented but... The doc states

plot_metrics is basically a seamless wrapper around clip_roi, cloud_metrics, cbind and adequate processing settings.

Then you look at clip_roi

... in clip_roi: optional supplementary options (see supported geometries). Unused in other functions

So we look at supported geometries

Polygon, SpatialPolygons, SpatialPolygonsDataFrame, SpatialPoints or SpatialPointsDataFrame in that case a parameter radius must be passed in ...

Then we look at radius from clip_circle

radius numeric. disc radius or radii.

As a conclusion, and to be honnest, it is not documented...

[benice2me]

Thank you so much for answering my questions. Although I really should have looked into the clip_roi documentation, you might want to add a little footnote about this to the doc in due time 😉

[benice2me]

I hate asking about this again but I am still unsure about the metrics which were calculated. I have tried the following 4 lines:

plot_metrics(catalog, .stdmetrics_z, plots, radius = plots$STRAAL) # the way you suggested, STRAAL is the header of my column containing radii
plot_metrics(catalog, .stdmetrics_z, plots, radius = 14) # 14 is the first radius in the list
plot_metrics(catalog, .stdmetrics_z, plots, radius = 10) # a somewhat arbitrary number
plot_metrics(catalog, .stdmetrics_z, plots, radius = list_r) # with list_r being a list of the radii

I have checked a couple of calculated metrics, such as zmax, zmean, zsd, zpcum7 and zpcum8, and they seem to be the exact same for every different radius input. Do you have an idea about how this is happening?

Sadly I cannot share my data with you.

[Jean-Romain]

This is not what I'm seeing with a reproducible example

LASfile <- system.file("extdata", "Megaplot.laz", package="lidR")
las = readLAScatalog(LASfile)
col = sf::sf.colors(15)

shpfile <- system.file("extdata", "efi_plot.shp", package="lidR")
inventory <- sf::st_read(shpfile, quiet = TRUE)
inventory # contains an ID and a Value Of Interest (VOI) per plot
fun = ~list(q85 = quantile(Z, probs = 0.85))

m1 <- plot_metrics(las, fun, inventory, radius = 11.28)
m2 <- plot_metrics(las, fun, inventory, radius = runif(5, 5, 25))

plot(m1["q85"], pch = 19, cex = 3)
plot(m2["q85"], pch = 19, cex = 3)

m1$q85
#> [1] 21.4225 22.0655 23.0420 19.0825 14.9125
m2$q85
#> [1] 21.000 21.680 22.890 21.541 14.660