lx_plot.py

#!/usr/bin/env python3

"""
"""

import os
import pandas as pd
import json
import plotly.graph_objs as go
from plotly.offline import plot
try:
    import utils
except ImportError:
    from templates import utils


__version__ = "0.0.1"
__build__ = "14.12.2020"
__template__ = "PROCESS_C90-nf"

logger = utils.get_logger(__file__)

if __file__.endswith(".command.sh"):
    LX_FILES = '$lx_files '.split()
    L_TARGET = float("$params.l_target")
    SCALE = '$scale'
    logger.debug("Running {} with parameters:".format(
        os.path.basename(__file__)))
    logger.debug("C90_FILES: {}".format(LX_FILES))
    logger.debug("L_TARGET: {}".format(L_TARGET))
    logger.debug("SCALE: {}".format(SCALE))


def main(c90files, l_target, scale):

    df_Lx = pd.DataFrame(columns=['Sample', 'Reference', 'Assembler', 'Lx', 'nContigs'])
    #df_Lx = pd.concat((pd.read_csv(f) for f in c90files))
    for file_c90 in c90files:
        sample_name = os.path.basename(file_c90).split('_')[0]
        data = pd.read_csv(file_c90)
        data['Sample'] = sample_name
        df_Lx = pd.concat([df_Lx, data], ignore_index=True)

    #df_Lx = df_Lx.astype({'nContigs': 'int32'})
    #print(df_Lx)

    # Create plot - Lx per reference for each sample
    report_dict = {}
    for sample in sorted(df_Lx['Sample'].unique()):
        for reference in sorted(df_Lx['Reference'].unique()):
            fig_Lx = go.Figure()
            i = 0
            for assembler in sorted(df_Lx['Assembler'].unique(), key=lambda v: v.upper()):
                if df_Lx['nContigs'][(df_Lx['Sample'] == sample) & (df_Lx['Reference'] == reference) & (df_Lx['Assembler'] == assembler)].nunique() > 1:
                    fig_Lx.add_trace(go.Scatter(x=df_Lx['Lx'][(df_Lx['Sample'] == sample) &
                                                            (df_Lx['Reference'] == reference) &
                                                            (df_Lx['Assembler'] == assembler)],
                                                y=df_Lx['nContigs'][(df_Lx['Sample'] == sample) &
                                                                    (df_Lx['Reference'] == reference) &
                                                                    (df_Lx['Assembler'] == assembler)],
                                                name=assembler, line=dict(color=utils.COLOURS[i], width=2)))
                    i += 1
            # add target line
            fig_Lx.add_shape(type="line", yref="paper",
                             x0=l_target*100, y0=0, x1=l_target*100, y1=1,
                             line=dict(color="#D3D3D3", width=4,dash="dashdot"))

            fig_Lx.update_layout(xaxis_title="L(x) %",
                                 yaxis_title='Contigs',
                                 plot_bgcolor='rgb(255,255,255)',
                                 xaxis=dict(showline=True, zeroline=False, linewidth=1, linecolor='black',
                                            gridcolor='#DCDCDC'))
            fig_Lx.update_xaxes(type=scale)

            plot(fig_Lx, filename='{0}_{1}_lx.html'.format(sample, reference.replace(' ', '_')), auto_open=False)
            plot_species = fig_Lx.to_json()

            if sample not in report_dict.keys():
                report_dict[sample] = {"PlotData": {reference: [plot_species]}}
            else:
                if reference not in report_dict[sample]["PlotData"].keys():
                    report_dict[sample]["PlotData"][reference] = [plot_species]
                else:
                    report_dict[sample]["PlotData"][reference].append(plot_species)


    with open("lx.json", "w") as json_report:
        json_report.write(json.dumps(report_dict, separators=(",", ":")))


if __name__ == '__main__':
    main(LX_FILES, L_TARGET, SCALE)