internship/olaflow/processing/diff.py

import argparse
import gzip
from itertools import starmap
import logging
from multiprocessing import pool
import pathlib
import pickle
import sys

from cycler import cycler
import matplotlib.pyplot as plt
import matplotlib.gridspec as gridspec
import numpy as np
from scipy import interpolate

from .olaflow import OFModel

parser = argparse.ArgumentParser(description="Post-process olaflow results")
parser.add_argument("-v", "--verbose", action="count", default=0)
parser.add_argument(
    "-o",
    "--output",
    action="append",
    type=pathlib.Path,
    help="Post-processing directory",
    required=True,
)
parser.add_argument(
    "-t",
    "--timestep",
    type=float,
    help="Time-step to compare",
)
parser.add_argument(
    "-f",
    "--func",
    type=str,
    help="Post-process function to compare",
    default="graphUniform",
    choices=("graphUniform", "graphUniform2"),
)
parser.add_argument(
    "-y",
    "--field",
    type=str,
    help="Field to compare",
    default="alpha.water",
    choices=("alpha.water", "U"),
)

args = parser.parse_args()

logging.basicConfig(level=max((10, 20 - 10 * args.verbose)))
log = logging.getLogger("ola_post")

log.info("Plotting comparison of model output")


def get_pickle(out):
    with (
        path.open("rb")
        if (path := out.joinpath("pickle")).exists()
        else gzip.open(path.with_suffix(".gz"), "rb")
    ) as f:
        return pickle.load(f)


models = list(map(get_pickle, args.output))


fig, ax_ = plt.subplots(
    len(models),
    figsize=(6, 1.5 * len(models)),
    dpi=100,
    constrained_layout=True,
    squeeze=False,
)
ax = ax_[:, 0]

if args.timestep is None:
    match args.field:
        case "alpha.water":
            for i, (_ax, _model) in enumerate(zip(ax, models)):
                _ax.contour(
                    _model.t,
                    _model.post_fields[args.func][f"x_{args.field}"],
                    _model.post_fields[args.func][args.field].T,
                    (0.5,),
                    colors="k",
                )
        case "U":
            for i, (_ax, _model) in enumerate(zip(ax, models)):
                _c = _ax.imshow(
                    np.where(_model.post_fields[args.func]["alpha.water"] > 0.5, np.linalg.norm(_model.post_fields[args.func][args.field], axis=2), np.nan).T[::-1],
                    vmin=0,
                    vmax=20,
                    cmap="inferno_r",
                    extent=(
                        _model.t.min(),
                        _model.t.max(),
                        _model.post_fields[args.func][f"x_{args.field}"].min(),
                        _model.post_fields[args.func][f"x_{args.field}"].max(),
                    ),
                )
                fig.colorbar(_c, label=f"{args.field} (m/s)", ax=_ax)
        case _:
            log.error(f"Cannot plot field {args.field} from {args.func}")
            sys.exit(1)

    for i, (_ax, _model) in enumerate(zip(ax, models)):
        _ax.set(xlabel="t (s)", ylabel="z (m)", title=f"Case {i}")
        _ax.grid(color="k", alpha=0.2)

    fig.savefig(
        args.output[0].joinpath(
            f"diff_{args.func}_{args.field}_{'_'.join([o.name for o in args.output])}.pdf"
        )
    )
else:
    match args.field:
        case "alpha.water":
            for i, (_ax, _model) in enumerate(zip(ax, models)):
                _ax.tricontour(
                    _model.x,
                    _model.z,
                    _model.fields[args.field][np.where(_model.t == args.timestep)[0]][0],
                    levels=(0.5,),
                    colors="k",
                )
        case _:
            log.error(f"Cannot plot field {args.field} from {args.func} at timestep")
            sys.exit(1)

    for i, (_ax, _model) in enumerate(zip(ax, models)):
        _ax.set(xlabel="x (m)", ylabel="z (m)", title=f"Case {i}")
        _ax.grid()

    fig.savefig(
        args.output[0].joinpath(
            f"diff_t{args.timestep}_{'_'.join([o.name for o in args.output])}.pdf"
        )
    )