"""
=================
Spectrum Database
=================

RADIS has :py:class:`~radis.tools.database.SpecDatabase` feature used to store
and retrieve calculated Spectrums. This feature is convinient if a spectrum takes a
long time to compute. A path can be specified for SpecDatabase all
Spectrums are stored as .spec files which can be loaded
from the SpecDatabase object itself.  A csv file is generated which contains all
input and conditional parameters of Spectrum.

RADIS also has :py:meth:`~radis.lbl.loader.DatabankLoader.init_database` feature
which initializes the SpecDatabase for the SpectrumFactory and every Spectrum
generated from it will be stored in the SpecDatabase automatically.

You can use :py:meth:`~radis.tools.database.SpecList.plot_cond` to make a 2D plot
using the conditions of the Spectrums in the SpecDatabase and use a z_label to plot
a heat map based on it.

:py:class:`~radis.tools.database.SpecDatabase` also has useful
:py:meth:`~radis.tools.database.SpecDatabase.fit_spectrum` and
:py:meth:`~radis.tools.database.SpecDatabase.interpolate` methods.


"""

from pathlib import Path

import numpy as np

from radis import SpectrumFactory
from radis.tools import SpecDatabase

sf = SpectrumFactory(
    wavenum_min=2900,
    wavenum_max=3200,
    molecule="OH",
    truncation=5,  # cm-1
    medium="vacuum",
    verbose=0,  # more for more details
    pressure=10,
    wstep=0.1,
)
sf.warnings = {"AccuracyError": "ignore"}
sf.fetch_databank("hitran")

# Generating a Spectrum
s1 = sf.eq_spectrum(Tgas=300, path_length=1)

# %%
# Creating SpecDatabase
# This is simply a local folder on your computer.
from radis.test.utils import getTestFile

my_folder = Path(getTestFile(".")) / "SpecDatabase_Test"
db = SpecDatabase(my_folder)

# %%
# There are many ways to add spectra to the database :
# Method 1: Creating .spec file and adding manually to SpecDatabase

db.add(s1)

# %%
# Method 2: Using init_database()
# SpecDatabase is connected to the SpectrumFactory ``sf`` above
# Generated Spectrum are added to SpecDatabase automatically :
sf.init_database(my_folder)

wstep = np.linspace(0.1, 0.001, 4)
Tgas_arr = np.linspace(300, 700, 7)


# Multiple Spectrum calculation based on different values of Tgas and wstep
for i in wstep:
    sf._wstep = i
    sf.params.wstep = i
    for Tgas in Tgas_arr:
        sf.eq_spectrum(Tgas=Tgas, path_length=1)


# Loading SpecDatabase
db_new = SpecDatabase(my_folder)

# Plot Tgas vs wstep for all Spectrums, heatmap based on calculation_time
db_new.plot_cond("Tgas", "wstep", "calculation_time")