Previous version performed "major pigment" checks using a set of hard-coded taxa-pigment pairs. When a major pigment was not included in the sample set, the taxa was dropped.
For phytoclass>=3.0.0, these major pigments are indicated using '2's in the F matrix.
The default behavior, specified by the following matrix:
check_mat <-
matrix(
c(
c(
"Chlorophytes", "Prasinophytes", "Prasinophytes", "Dinoflagellates-1",
"Diatoms-1", "Diatoms-2", "Syn", "Cryptophytes", "Haptophytes-H",
"Haptophytes-L", "Diatoms-1", "Pelagophytes", "Prasinophytes"
),
c(
"Chl_b", "Chl_b", "Pra", "Per", "Chl_c1", "Fuco", "Zea", "Allo",
"X19hex", "X19hex", "Fuco", "X19but", "Chl_b"
)
),
ncol = 2
)
Is reproduced using an updated default F matrix that includes '2's.
Some very rare edge cases will function differently. Older custom F matrices using only '1's will no longer do the major pigment check for the exact string matches above.
min_max.rda using new pigment ratio min/max from Simon Wright.min_max.This version update removes internal functions that became unused once combined into a single function. The sets of functions include "Conduit", "Fac_F_RR", "Minimise_elements" and "Random_neighbor".
convergence_figure() was added in simulated_annealing_Prochloro()simulated_annealing_Prochloro()Cluster() now accepts a row_ids input vector to name each sampleCluster() exports an assigments item that relates the ID to the clusterCluster() checks if data.frame input has a column of strings and removes
them prior to clusteringerror if the number of colors needed exceeded 11.
Now, the figure functions can handle additional colors.Cluster() had an issue if given a matrix without rownames and would error,
now will convert to data.frame prior to running boxcar normalisingWe have added a function which allows for chlorophyll derivations for prochlorococcus.
We have improved documentation, and relabelled groups, such as 'Diatom-1' instead of 'Diatom-A'.
NEWS.md file to track changes to the package.