Subsampling_snakefile

import os

data_folder = config["data_folder"]
num_samples= config["num_samples"]

if not os.path.exists(data_folder+"/selected_data/"):
    os.makedirs(data_folder+"/selected_data/")

rule all:
    input:
        data_folder+"/convert_input_to_fasta.done"


def get_subdirs(data_folder):
    return [f.path for f in os.scandir(data_folder) if f.is_dir() and "plot" not in f.path and "benchmarking" not in f.path]


# save num_taxa and num_sequences for every nexus or fasta file in a subdirectory of data_folder
rule get_nexus_data:
    output:
        touch(data_folder+"/get_nexus_data.done"),
        data_folder+"/nexus_data.csv"
    params:
        data=data_folder
    shell:
        "./scripts/get_nexus_data.sh  {params.data}"


rule select_datasets:
    input:
        data_folder+"/get_nexus_data.done",
        data_csv=data_folder+"/nexus_data.csv"
    params:
        selected_datasets_csv=data_folder+"/selected_datasets.csv",
        num_samples=num_samples,
        data=data_folder,
    output:
        temp(touch(data_folder+"/select_datasets.done")),
    script:
        "scripts/select_datasets.py"


# convert input alignments from nexus to fasta, if necessary
rule convert_input_to_fasta:
    input:
        data_folder+"/select_datasets.done"
    output:
        temp(touch(data_folder+"/convert_input_to_fasta.done"))
    params:
        data_folder = data_folder+"/selected_data/",
    script:
        "scripts/convert_input_to_fasta.py"


# rule run_stability_analysis:
#     input:
#         data_folder+"/convert_input_to_fasta.done"
#     output:
#         data_folder+"/selected_data/plots/random_forest_results.pdf",
#         data_folder+"/selected_data/plots/p_au_proportion_plot.pdf"
#     shell:
#         '''
#         snakemake --unlock
#         # snakemake -c36 --touch --rerun-incomplete
#         snakemake -c32 -R random_forest_regression
#         '''