Let’s use a small sample of a real dataset with data on Australian
-native plant species including Eucalypts and Acacias. The dataset
-contains lots of measurement fields attached to each species occurrence
-that contain data the organism’s traits or environment. I have extracted
-3 examples: LMA_g.m2
, LeafN_area_g.m2
,
-PNUE
.
+Let’s use a small sample of a real dataset of Australian native plant
+species, including Eucalypts and Acacias. The dataset contains lots of
+columns containing measurements that capture the traits or environment
+of each species occurrence. I have extracted 3 examples of these
+measurement columns: LMA_g.m2
,
+LeafN_area_g.m2
, PNUE
.
Darwin Core confusingly handles measurement fields. This nests them
-and assigns them to the correct columns. You just specify the column,
-the unit, and the type of measurement
+The way that Darwin Core handles measurement fields is slightly
+different to how they are organised in these data. In
+df_filtered
, lots of information about what data the column
+contains is in the abbreviated column name. This includes the type of
+measure, the unit it is measured in, and the value itself. In Darwin
+Core, this information needs to be documented individually for each
+measurement, which requires multiple columns and multiple rows (ie a
+‘long’ data format).
+Then you can unnest that column, and it will have all the correct
-columns formatted correctly.
+
df_nested |>
unnest(measurementOrFact)
diff --git a/authors.html b/authors.html
index 366c02f..1c3774e 100644
--- a/authors.html
+++ b/authors.html
@@ -3,7 +3,7 @@
Skip to contents
-