Preparing the S Matrix

To run phytoclass you must prepare a sample matrix (aka the S matrix). For this we recommend creating a comma-separated-value (.csv) file. Several example files are included in the ./vignettes/ directory of the phytoclass source code. .csv files can be created in any plain text editor. Below is an example S-matrix .csv file content:

"","Per","X19but","Fuco","Neox","Pra","Viol","X19hex","Allo","Zea","Lut","ChlcMGDG18","ChlcMGDG14","Chl_b","Tchla"
"1",0,0.03024,0.06225,0.00557,0.01407,0.00759,0.08224,0.00188,0.00201,0,0,0,0.08661,0.45851
"2",0,0.01084,0.02864,0.00111,0.00351,0.00144,0.01497,0.00144,0.00191,0,0,0,0.01473,0.14571
"3",0,0.0156,0.2172,0.0064,0.0092,0,0.0174,0.0058,0.0036,0,0,0,0.057,0.6127
"4",0,0.0177,0.2347,0.007,0.0115,0,0.0189,0.0054,0.0038,0,0,0,0.0619,0.6207
"5",0,0.0252,0.2952,0.0099,0.013,0,0.0211,0.0014,0.0076,0,0,0,0.0578,0.5302
"6",0.0102,0.0222,0.2275,0.0076,0.0107,0,0.019,0,0.003,0,0,0,0.0453,0.4057
"7",0,0.0151,0.1417,0.004,0.008,0.0026,0.0255,8e-04,0.001,0,0,0,0.0312,0.2117
"8",0,0.0452,0.4484,0.0037,0.0065,0,0.1317,0.0044,0.0023,0,0,0,0.0093,0.6885
"9",0,0.0518,0.6228,0.0049,0.0075,0.004,0.1321,0.0597,0.0035,0,0,0,0.0121,1.0849
"10",0,0.022,0.109,0.004,0.008,0.003,0.042,0.003,0.003,0,0,0,0.032,0.206
"11",0,0.049062693,0.526641835,0.017334346,0.034463572,0,0.120442872,0.029990012,0.039966194,0,0,0,0.153250497,1.039660729
"12",0,0.020526621,0.271493439,0.017692288,0.019947584,0,0.030115477,0,0.030828916,0,0,0,0.09243326,0.481204323
"13",0,0.022161768,0.165425707,0.012242414,0.019562855,0,0.027210489,0,0.017661542,0,0,0,0.087205952,0.340907892
"14",0,0.014385454,0.255559406,0.020671983,0.047363662,0.011507712,0.021676506,0.009032764,0.018938989,0,0,0,0.142134419,0.489290176

This example includes only 14 samples (rows), with the pigment concentration of each sample included on each row. 14 is the minumum number of rows required to apply simulated annealing. The first row includes the names for each pigment. These pigment abbreviations must match exactly to the pigments in the F-matrix. For this example we are using the default phytoclass:Fm F matrix, and the column names in the example above should be used.

NOTE: Instead of using a plain text editor, .csv files can also be created using a spreadsheet editor (MS Excel, Google Docs, LibreOffice Calc, etc.) by using “save as” or “export” to .csv.

Simulated Annealing

To run the simulated annealing call the function with your S matrix passed as the only argument.

# load the sample matrix from a file
S_matrix <- read.csv("vignettes/custom-example-S.csv")

# run simulated annealing 
results <- phytoclass::simulated_annealing(S_matrix)

Viewing Results

Results of the simulated annealing run are in the results object returned by the function.

results$`condition number`
results$RMSE
results$MAE
results$Error
results$`Class abundances`
results$Figure