INNER CODE UNIT · Python

get_taxa_cols

hahnlab/CAFE5 · scripts/remove_taxa.py:3

def get_taxa_cols(header_order_dict, taxa_list):
    taxa_cols = [header_order_dict[taxon_name] for taxon_name in taxa_list]
    return taxa_cols
    
def remove_taxa(input_file_path, output_file_path, taxa_to_remove):
    taxa_list = taxa_to_remove.split(',')
    header_order = dict() # node name:idx
    node_name_idx = 0

    with open(output_file_path, 'w') as output_file:
        with open(input_file_path, 'r') as input_file:
            for line in input_file:
                line = line.rstrip()

                if line.startswith('#'):
                    node_name = line.lstrip('#')

                    if not node_name in taxa_list:

View source record →