Skip to content

updating SDM tutorial #13

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
wants to merge 8 commits into
base: master
Choose a base branch
from
Open

updating SDM tutorial #13

Changes from all commits
Commits
Show all changes
8 commits
Select commit Hold shift + click to select a range
2a9a879
sp_occurrence update
Jul 27, 2023
65cba36
Extended map boundaries to show all data
Jul 27, 2023
293beac
minor typos
Jul 27, 2023
71065fe
Corrected terra::extract argument order
Jul 27, 2023
4dcdbfd
Corrected reference to wrld_simpl
Jul 27, 2023
1f71005
Added the snap = "out" argument to extend
Jul 27, 2023
5866100
Added note for the source of `gridSample`
Jul 27, 2023
e4b2436
updated exercises
Jul 27, 2023
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
32 changes: 16 additions & 16 deletions source/sdm/_R/2_sdm_occdata.rmd
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -36,16 +36,16 @@ If you do not have any species distribution data you can get started by download


```{r, sdm11B, eval=FALSE}
acaule <- geodata::sp_occurrence("solanum", "acaule*", geo=FALSE)
acaule <- geodata::sp_occurrence("solanum", "acaule", geo=FALSE)
## Loading required namespace: jsonlite
## 7238 records found
## 0-300-600-900-1200-1500-1800-2100-2400-2700-3000-3300-3600-3900-4200
## 6974 records found
## 0-300-600-900-1200-1500-1800-2100-2400-2700-3000-3300-3600-3900-4200-4500-4800-5100-5400-5700-
## 6000-6300-6600-6900-6974
```

If you want to understand the order of the arguments given here to `gbif` or find out what other arguments you can use with this function, check out the help file (remember you can't access help files if the library is not loaded), by typing: `?gbif` or `help(gbif)`. Note the use of the asterix in "acaule*" to not only request `Solanum acaule`, but also variations such as the full name, *Solanum acaule* Bitter, or subspecies such as *Solanum acaule* subsp. *aemulans*.

If you want to understand the order of the arguments given here to `sp_occurrence` or find out what other arguments you can use with this function, check out the help file (remember you can't access help files if the library, in this case `geodata`, is not loaded), by typing: `?sp_occurrence` or `help(sp_occurrence)`.

Many occurrence records may not have geographic coordinates. In this case, out of the 1366 records that GBIF returned (January 2013), there were 1082 records with coordinates,
Many occurrence records may not have geographic coordinates. In this case, out of the 1366 records that GBIF returned (in January 2013), there were 1082 records with coordinates,

```{r, sdm2}
# load the saved S. acaule data
Expand All @@ -69,7 +69,7 @@ Below is a simple way to make a map of the occurrence localities of *Solanum aca
```{r, sdm3}
library(geodata)
wrld <- world(path=".")
plot(wrld, xlim=c(-110,60), ylim=c(-80,40), col="light yellow", border="light gray")
plot(wrld, xlim=c(-110,90), ylim=c(-80,40), col="light yellow", border="light gray")
# add the points
points(acgeo$lon, acgeo$lat, col='red', pch=20)
```
Expand All @@ -79,11 +79,11 @@ The `wrld` dataset contains rough country outlines. You can use other datasets o

## Data cleaning

Data 'cleaning' is particularly important for data sourced from species distribution data warehouses such as GBIF. Such efforts do not specifically gather data for the purpose of species distribution modeling, so you need to understand the data and clean them appropriately, for your application. Here we provide an example.
Data 'cleaning' is particularly important for data sourced from species distribution data warehouses such as GBIF. Such efforts do not specifically gather data for the purpose of species distribution modeling, so you need to understand the data and clean them appropriately, for your application. Here we provide an example.

`Solanum acaule` is a species that occurs in the higher parts of the Andes mountains of southern Peru, Bolivia and northern Argentina. Do you see any errors on the map?

There are a few records that map in the ocean just south of Pakistan. Any idea why that may have happened? It is a common mistake, missing minus signs. The coordinates are around (65.4, 23.4) but they should in Northern Argentina, around (-65.4, -23.4) (you can use the "click" function to query the coordintates on the map). There are two records (rows 303 and 885) that map to the same spot in Antarctica (-76.3, -76.3). The locality description says that is should be in Huarochiri, near Lima, Peru. So the longitude is probably correct, and erroneously copied to the latitude. Interestingly the record occurs twice. The orignal source is the International Potato Center, and a copy is provided by "SINGER" that aling the way appears to have "corrected" the country to Antarctica:
There are a few records that map in the ocean just south of Pakistan. Any idea why that may have happened? It is a common mistake, missing minus signs. The coordinates are around (65.4, 23.4) but they should in Northern Argentina, around (-65.4, -23.4) (you can use the "click" function to query the coordintates on the map). There are two records (rows 303 and 885) that map to the same spot in Antarctica (-76.3, -76.3). The locality description says that is should be in Huarochiri, near Lima, Peru. So the longitude is probably correct, and erroneously copied to the latitude. Interestingly the record occurs twice. The orignal source is the International Potato Center, and a copy is provided by "SINGER" that along the way appears to have "corrected" the country to Antarctica:

```{r, sdm4a}
acgeo[c(303,885),1:10]
Expand Down Expand Up @@ -153,10 +153,10 @@ class(acv)
We can now use do a spatial query of the polygons in `wrld`

```{r, sdm6b}
ovr <- extract(acv, wrld)
ovr <- extract(wrld, acv)
```

Object 'ovr' has, for each point, the matching record from `wrld`. We need the variable 'NAME_0' in the data.frame of wrld_simpl
Object 'ovr' has, for each point, the matching record from `wrld`. We need the variable 'NAME_0' in the data.frame of `wrld`.


```{r, sdm6c}
Expand All @@ -176,7 +176,7 @@ colnames(m) <- c("polygons", "acaule")
m
```

In this case the mismatch is probably because wrld_simpl is not very precise as the records map to locations very close to the border between Bolivia and its neighbors.
In this case the mismatch is probably because `wrld` is not very precise as the records map to locations very close to the border between Bolivia and its neighbors.

```{r, sdm6e}
plot(acv)
Expand Down Expand Up @@ -222,7 +222,7 @@ r <- rast(acv)
res(r) <- 1

# extend (expand) the extent of the SpatRaster a little
r <- extend(r, ext(r)+1)
r <- extend(r, ext(r)+1, snap = "out")

# sample:
set.seed(13)
Expand All @@ -237,7 +237,7 @@ points(acsel, cex=1, col='red', pch='x')
```


Note that with the `gridSample` function you can also do 'chess-board' sampling. This can be useful to split the data in 'training' and 'testing' sets (see the model evaluation chapter).
Note that with the `gridSample` function from the `dismo` package you can also do 'chess-board' sampling. This can be useful to split the data in 'training' and 'testing' sets (see the model evaluation chapter).

At this point, it could be useful to save the cleaned data set. For example, you can use `as.data.frame(acsel)` and then `write.csv`. Or you can use `pack` and `saveRDS` so that we can use them later. We did that, and the saved file is available from the `predicts` package and can be read like this:

Expand All @@ -251,7 +251,7 @@ In a real research project you would want to spend much more time on this first

## 2.8 Exercises

1. Use the gbif function to download records for the African elephant (or another species of your preference, try to get one with between 10 and 100 records). Use option "geo=FALSE" to also get records with no (numerical) georeference.
1. Use the `sp_occurrence` function to download records for the African elephant (or another species of your preference, try to get one with between 10 and 100 records). Use option "geo=FALSE" to also get records with no (numerical) georeference.

2. Summarize the data: how many records are there, how many have coordinates, how many records without coordinates have a textual georeference (locality description)?

Expand All @@ -263,7 +263,7 @@ In a real research project you would want to spend much more time on this first

More advanced:

6. Use the 'rasterize' function to create a raster of the number of observations and make a map. Use "wrld_simpl" from the maptools package for country boundaries.
6. Use the 'rasterize' function to create a raster of the number of observations and make a map. Use "world" from the geodata package for country boundaries.

7. Map the uncertainty associated with the georeferences. Some records in data returned by gbif have that. You can also extract it from the data returned by the geocode function.