Check against hard-coded pigment names is fragile

[7yl4r]

The following check will not work if unusual pigment names are used.

phytoclass-shiny-gui/app/server.R

Lines 55 to 87 in 1ea5610

	# ---- Matrix normalization helpers ----
	# Uploaded files may include ID columns instead of rownames.
	# These helpers standardize matrices before rendering.
	normalize_samples <- function(df){
	first_name <- colnames(df)[1]
	# If first column does not look like a pigment, treat it as sample IDs
	pigment_like <- grepl("Per|Fuco|Pra|Zea|Tchla|Chl|X19", first_name)
	if(!pigment_like){
	ids <- make.unique(as.character(df[[1]]))
	rownames(df) <- ids
	df <- df[, -1, drop = FALSE]
	}
	df
	}
	normalize_taxa <- function(df){
	if(ncol(df) == 0) return(df)
	first_name <- colnames(df)[1]
	# Same logic as samples: detect whether the first column holds taxa names
	pigment_like <- grepl("Per|Fuco|Pra|Zea|Tchla|Chl|X19", first_name)
	if(!pigment_like){
	ids <- make.unique(as.character(df[[1]]))
	rownames(df) <- ids
	df <- df[, -1, drop = FALSE]
	}
	df
	}

Previously this was a check if the column was characters.

---

I wonder if these checks are entirely necessary. We want the phytoclass library to be handling warnings and error checking rather than the GUI.