AeroViz 0.1.3__py3-none-any.whl → 0.1.3b0__py3-none-any.whl

AeroViz/plot/improve/improve.py

@@ -1,240 +0,0 @@
-import matplotlib.pyplot as plt
-import numpy as np
-import pandas as pd
-from matplotlib.pyplot import Figure, Axes
-from pandas import DataFrame, read_json
-from scipy.optimize import curve_fit
-from pathlib import Path
-
-from AeroViz import plot
-from AeroViz.plot.utils import *
-from AeroViz.tools import DataBase, DataReader, DataClassifier
-
-# TODO: this file has to be reorganized
-
-__all__ = ['chemical_enhancement',
-		   'ammonium_rich',
-		   'pie_IMPROVE',
-		   'MLR_IMPROVE',
-		   'fRH_plot',
-		   ]
-
-
-@set_figure
-def chemical_enhancement(data_set: DataFrame = None,
-						 data_std: DataFrame = None,
-						 ax: Axes | None = None,
-						 **kwargs
-						 ) -> tuple[Figure, Axes]:
-	fig, ax = plt.subplots() if ax is None else (ax.get_figure(), ax)
-
-	ser_grp_sta, ser_grp_sta_std = DataClassifier(DataBase('/Users/chanchihyu/NTU/2020能見度計畫/data/All_data.csv'),
-												  by='State')
-	species = ['AS', 'AN', 'POC', 'SOC', 'Soil', 'SS', 'EC', 'ALWC']
-	data_set, data_std = ser_grp_sta.loc[:, species], ser_grp_sta_std.loc[:, species]
-
-	width = 0.20
-	block = width / 4
-
-	x = np.array([1, 2, 3, 4, 5, 6, 7])
-	for i, state in enumerate(['Clean', 'Transition', 'Event']):
-		val = np.array(data_set.iloc[i, :-1])
-		std = (0,) * 7, np.array(data_std.iloc[i, :-1])
-
-		plt.bar(x + (i + 1) * (width + block), val, yerr=std, width=width, color=Color.colors3[:-1],
-				alpha=0.6 + (0.2 * i),
-				edgecolor=None, capsize=None, label=state)
-
-	ax.set(xlabel=r'$\bf Chemical\ species$',
-		   ylabel=r'$\bf Mass\ concentration\ ({\mu}g/m^3)$',
-		   xticks=x + 2 * (width + block),
-		   xticklabels=species,
-		   ylim=(0, 25),
-		   title=r'$\bf Chemical\ enhancement$')
-
-	ax.vlines(8, 0, 25, linestyles='--', colors='k')
-
-	ax2 = ax.twinx()
-	for i, state in enumerate(['Clean', 'Transition', 'Event']):
-		val = np.array(data_set.iloc[i, -1])
-		std = np.array([[0], [data_std.iloc[i, -1]]])
-		plt.bar(8 + (i + 1) * (width + block), val, yerr=std, width=width, color='#96c8e6',
-				alpha=0.6 + (0.2 * i), edgecolor=None, capsize=None, label=state)
-
-	ax2.set(ylabel=r'$\bf Mass\ concentration\ ({\mu}g/m^3)$',
-			ylim=(0, 100),
-			xticks=x + 2 * (width + block),
-			xticklabels=species
-			)
-
-	a = (np.array(data_set.loc['Event']) + np.array(data_set.loc['Transition'])) / 2
-	b = (np.array(data_set.loc['Transition']) + np.array(data_set.loc['Clean'])) / 2
-	c = np.array(data_set.loc['Event']) / np.array(data_set.loc['Transition'])
-	d = np.array(data_set.loc['Transition']) / np.array(data_set.loc['Clean'])
-
-	for i, (posa, posb, vala, valb) in enumerate(zip(a, b, c, d)):
-		if i < 7:
-			ax.text(i + 1.5, posa, '{:.2f}'.format(vala), fontsize=6, weight='bold', zorder=1)
-			ax.text(i + 1.25, posb, '{:.2f}'.format(valb), fontsize=6, weight='bold', zorder=1)
-		else:
-			ax2.text(i + 1.5, posa, '{:.2f}'.format(vala), fontsize=6, weight='bold', zorder=1)
-			ax2.text(i + 1.25, posb, '{:.2f}'.format(valb), fontsize=6, weight='bold', zorder=1)
-
-	plt.show()
-
-	return fig, ax
-
-
-@set_figure
-def ammonium_rich(df: DataFrame,
-				  **kwargs
-				  ) -> tuple[Figure, Axes]:
-	df = df[['NH4+', 'SO42-', 'NO3-', 'PM25']].dropna().copy().div([18, 96, 62, 1])
-	df['required_ammonium'] = df['NO3-'] + 2 * df['SO42-']
-
-	fig, ax = plt.subplots()
-
-	scatter = ax.scatter(df['required_ammonium'].to_numpy(), df['NH4+'].to_numpy(), c=df['PM25'].to_numpy(),
-						 vmin=0, vmax=70, cmap='jet', marker='o', s=10, alpha=1)
-
-	ax.axline((0, 0), slope=1., color='k', lw=2, ls='--', alpha=0.5, label='1:1')
-	plt.text(0.97, 0.97, r'$\bf 1:1\ Line$', color='k', ha='right', va='top', transform=ax.transAxes)
-
-	ax.set(xlim=(0, 1.2),
-		   ylim=(0, 1.2),
-		   xlabel=r'$\bf NO_{3}^{-}\ +\ 2\ \times\ SO_{4}^{2-}\ (mole\ m^{-3})$',
-		   ylabel=r'$\bf NH_{4}^{+}\ (mole\ m^{-3})$',
-		   title=kwargs.get('title', ''))
-
-	color_bar = plt.colorbar(scatter, label=Unit('PM25'), extend='both')
-
-	# fig.savefig(f'Ammonium_rich_{title}')
-	plt.show()
-
-	return fig, ax
-
-
-def pie_IMPROVE():
-	Species1 = ['AS_ext_dry', 'AN_ext_dry', 'OM_ext_dry', 'Soil_ext_dry', 'SS_ext_dry', 'EC_ext_dry']
-	Species2 = ['AS_ext_dry', 'AN_ext_dry', 'OM_ext_dry', 'Soil_ext_dry', 'SS_ext_dry', 'EC_ext_dry', 'ALWC_ext']
-	Species3 = ['AS_ext', 'AN_ext', 'OM_ext', 'Soil_ext', 'SS_ext', 'EC_ext']
-
-	ser_grp_sta, _ = DataClassifier(DataBase(), by='State')
-
-	ext_dry_dict = ser_grp_sta.loc[:, Species1]
-	ext_amb_dict = ser_grp_sta.loc[:, Species2]
-	ext_mix_dict = ser_grp_sta.loc[:, Species3]
-
-	plot.donuts(data_set=ext_dry_dict, labels=['AS', 'AN', 'OM', 'Soil', 'SS', 'BC'], unit='Extinction')
-	plot.donuts(data_set=ext_mix_dict, labels=['AS', 'AN', 'OM', 'Soil', 'SS', 'BC'], unit='Extinction')
-	plot.donuts(data_set=ext_amb_dict, labels=['AS', 'AN', 'OM', 'Soil', 'SS', 'BC', 'ALWC'],
-				unit='Extinction', colors=Color.colors2)
-
-
-def MLR_IMPROVE(**kwargs):
-	"""
-	Perform multiple linear regression analysis and generate plots based on IMPROVE dataset.
-
-	Parameters
-	----------
-	**kwargs : dict
-		Additional keyword arguments for customization.
-
-	Returns
-	-------
-	None
-
-	Examples
-	--------
-	Example usage of MLR_IMPROVE function:
-
-	>>> MLR_IMPROVE()
-
-	Notes
-	-----
-	This function performs multiple linear regression analysis on the IMPROVE dataset and generates plots for analysis.
-
-	- The function first selects specific species from the dataset and drops NaN values.
-	- It calculates a 'Localized' value based on a multiplier and the sum of selected species.
-	- Data from 'modified_IMPROVE.csv' and 'revised_IMPROVE.csv' are read and concatenated with the dataset.
-	- Statistical analysis is performed using DataClassifier to calculate mean and standard deviation.
-	- Plots are generated using linear_regression for Extinction vs. Revised/Modified/Localized and Pie.donuts for a
-	  pie chart showing the distribution of species based on Extinction.
-
-	"""
-	species = ['Extinction', 'Scattering', 'Absorption',
-			   'total_ext_dry', 'AS_ext_dry', 'AN_ext_dry', 'OM_ext_dry', 'Soil_ext_dry', 'SS_ext_dry', 'EC_ext_dry',
-			   'AS', 'AN', 'POC', 'SOC', 'Soil', 'SS', 'EC', 'OM']
-
-	df = DataBase('/Users/chanchihyu/NTU/2020能見度計畫/data/All_data.csv')[species].dropna().copy()
-
-	# multiple_linear_regression(df, x=['AS', 'AN', 'POC', 'SOC', 'Soil', 'SS'], y='Scattering', add_constant=True)
-	# multiple_linear_regression(df, x=['POC', 'SOC', 'EC'], y='Absorption', add_constant=True)
-	# multiple_linear_regression(df, x=['AS', 'AN', 'POC', 'SOC', 'Soil', 'SS', 'EC'], y='Extinction', add_constant=False)
-
-	multiplier = [2.675, 4.707, 11.6, 7.272, 0, 0.131, 10.638]
-	df['Localized'] = df[['AS', 'AN', 'POC', 'SOC', 'Soil', 'SS', 'EC']].mul(multiplier).sum(axis=1)
-	# TODO: remove name
-	modify_IMPROVE = DataReader('modified_IMPROVE.csv')['total_ext_dry'].rename('Modified')
-	revised_IMPROVE = DataReader('revised_IMPROVE.csv')['total_ext_dry'].rename('Revised')
-
-	df = pd.concat([df, revised_IMPROVE, modify_IMPROVE], axis=1)
-
-	n_df = df[['AS', 'AN', 'POC', 'SOC', 'Soil', 'SS', 'EC']].mul(multiplier)
-	mean, std = DataClassifier(n_df, 'State')
-
-	ser_grp_sta, _ = DataClassifier(DataBase(), by='State')
-	mass_comp = ser_grp_sta.loc[:, ['AS', 'AN', 'POC', 'SOC', 'Soil', 'SS', 'EC']]
-
-	# plot
-	plot.linear_regression(df, x='Extinction', y=['Revised', 'Modified', 'Localized'], xlim=[0, 400], ylim=[0, 400],
-						   regression=True, diagonal=True)
-	plot.donuts(data_set=mass_comp, labels=['AS', 'AN', 'POC', 'SOC', 'Soil', 'SS', 'EC'],
-				unit='PM25', colors=Color.colors3)
-	plot.donuts(mean, labels=['AS', 'AN', 'POC', 'SOC', 'Soil', 'SS', 'EC'], unit='Extinction', colors=Color.colors3)
-
-
-@set_figure
-def fRH_plot(**kwargs) -> tuple[Figure, Axes]:
-	frh = read_json(Path(__file__).parent.parent / 'utils' / 'fRH.json')
-
-	def fitting_func(RH, a, b, c):
-		f = a + b * (RH / 100) ** c
-		return f
-
-	x = frh.index.to_numpy()
-	y = frh['fRHs'].to_numpy()
-
-	result = curve_fit(fitting_func, x, y)
-	params = result[0].tolist()
-	val_fit = fitting_func(x, *params)
-
-	fig, ax = plt.subplots(figsize=(3, 3))
-
-	ax.plot(frh.index, frh['fRH'], 'k-o', ms=2, label='$f(RH)_{original}$')
-	ax.plot(frh.index, frh['fRHs'], 'g-o', ms=2, label='$f(RH)_{small\\ mode}$')
-	ax.plot(frh.index, frh['fRHl'], 'r-o', ms=2, label='$f(RH)_{large\\ mode}$')
-	ax.plot(frh.index, frh['fRHSS'], 'b-o', ms=2, label='$f(RH)_{sea\\ salt}$')
-
-	ax.set(xlim=(0, 100),
-		   ylim=(1, None),
-		   xlabel='$RH (\\%)$',
-		   ylabel='$f(RH)$',
-		   title=f'$Hygroscopic\\ growth\\ factor$'
-		   )
-
-	ax.grid(axis='y', color='gray', linestyle='dashed', linewidth=0.4, alpha=0.4)
-
-	ax.legend()
-
-	plt.show()
-	# fig.savefig('fRH_plot')
-
-	return fig, ax
-
-
-if __name__ == '__main__':
-	# chemical_enhancement()
-	# MLR_IMPROVE()
-	# ammonium_rich()
-	fRH_plot()

AeroViz/plot/optical/aethalometer.py

@@ -1,77 +0,0 @@
-import matplotlib.pyplot as plt
-import numpy as np
-from pandas import date_range
-
-from AeroViz.plot.utils import *
-
-__all__ = ['plot_MA350',
-		   'plot_MA3502',
-		   'plot_day_night']
-
-
-@set_figure(figsize=(15, 5))
-def plot_MA350(df, **kwargs):
-	fig, ax = plt.subplots()
-
-	# ax.scatter(df.index, df['UV BCc'], marker='o', c='purple', alpha=0.5, label='UV BCc')
-	# ax.scatter(df.index, df['Blue BCc'], c='b', alpha=0.5, label='Blue BCc')
-	# ax.scatter(df.index, df['Green BCc'], c='g', alpha=0.5, label='Green BCc')
-	# ax.scatter(df.index, df['Red BCc'], c='r', alpha=0.5, label='Red BCc')
-	mean, std = round(df.mean(), 2), round(df.std(), 2)
-
-	label1 = rf'$MA350-0171\ :\;{mean["MA350_0171 IR BCc"]}\;\pm\;{std["MA350_0171 IR BCc"]}\;(ng/m^3)$'
-	label2 = rf'$MA350-0176\ :\;{mean["MA350_0176 IR BCc"]}\;\pm\;{std["MA350_0176 IR BCc"]}\;(ng/m^3)$'
-	label3 = rf'$BC-1054\ :\;{mean["BC1054 IR BCc"]}\;\pm\;{std["BC1054 IR BCc"]}\;(ng/m^3)$'
-	ax.scatter(df.index, df['MA350_0171 IR BCc'], s=10, ls='-', marker='o', c='#a3b18a', alpha=0.5, label=label1)
-	ax.scatter(df.index, df['MA350_0176 IR BCc'], s=10, ls='-', marker='o', c='#3a5a40', alpha=0.5, label=label2)
-	ax.scatter(df.index, df['BC1054 IR BCc'], s=10, ls='-', marker='o', c='g', alpha=0.5, label=label3)
-	ax.legend(prop={'weight': 'bold'}, loc='upper left')
-
-	st_tm, fn_tm = df.index[0], df.index[-1]
-	tick_time = date_range(st_tm, fn_tm, freq=kwargs.get('freq', '10d'))
-
-	ax.set(xlabel=kwargs.get('xlabel', ''),
-		   ylabel=kwargs.get('ylabel', r'$BC\ (ng/m^3)$'),
-		   xticks=kwargs.get('xticks', tick_time),
-		   xticklabels=kwargs.get('xticklabels', [_tm.strftime("%F") for _tm in tick_time]),
-		   xlim=kwargs.get('xlim', (st_tm, fn_tm)),
-		   ylim=kwargs.get('ylim', (0, None)),
-		   )
-
-
-@set_figure
-def plot_MA3502(df):
-	fig, ax = plt.subplots()
-
-	bins = np.array([375, 470, 528, 625, 880])
-	vals = df.dropna().iloc[:, -5:].values
-
-	ax.boxplot(vals, positions=bins, widths=20,
-			   showfliers=False, showmeans=True, meanline=True, patch_artist=True,
-			   boxprops=dict(facecolor='#f2c872', alpha=.7),
-			   meanprops=dict(color='#000000', ls='none'),
-			   medianprops=dict(ls='-', color='#000000'))
-
-	ax.set(xlim=(355, 900),
-		   ylim=(0, None),
-		   xlabel=r'$\lambda\ (nm)$',
-		   ylabel=r'$Absorption\ (1/Mm)$', )
-
-
-@set_figure(figsize=(6, 5))
-def plot_day_night(df):
-	# Group by hour of day and calculate mean
-	df_grouped = df.groupby(df.index.hour).mean()
-
-	# Create figure and plot
-	fig, ax = plt.subplots()
-	ax.plot(df_grouped.index, df_grouped['MA350_0171 IR BCc'], marker='o', c='k', alpha=0.5, label='MA350-0171')
-	ax.plot(df_grouped.index, df_grouped['MA350_0176 IR BCc'], marker='o', c='r', alpha=0.5, label='MA350-0176')
-	ax.plot(df_grouped.index, df_grouped['BC1054 IR BCc'], marker='o', c='b', alpha=0.5, label='BC-1054')
-
-	ax.set(xlim=(0, 23),
-		   xlabel='Hour of Day',
-		   ylabel=r'$BC\ (ng/m^3)$',
-		   title=f'Diurnal pattern', )
-
-	ax.legend()

AeroViz/plot/templates/event_evolution.py

@@ -1,65 +0,0 @@
-from os.path import join as pth
-
-import matplotlib.pyplot as plt
-import numpy as np
-from pandas import date_range, read_csv
-
-from AeroViz.plot.utils import *
-
-# TODO:
-
-# read csv file
-blh = read_csv(pth('事件分析.csv'), parse_dates=['Time'], index_col='Time')
-
-
-@set_figure(figsize=(12, 5))
-def event_evolution(_df, **kwargs):
-	print(f'Plot : {_df.month[0]}')
-
-	st_tm, fn_tm = _df.index[0], _df.index[-1]
-	tick_time = date_range(st_tm, fn_tm, freq='1d')  # set tick
-
-	# seperate day and night
-	_df_day = _df.between_time('6:00', '17:00').reindex(date_range(st_tm, fn_tm, freq='1h'))
-	_df_night = _df.between_time('18:00', '5:00').reindex(date_range(st_tm, fn_tm, freq='1h'))
-
-	## plot
-	fig, ax = plt.subplots()
-
-	## plot background
-	shade_value, _ = np.meshgrid(_df['PM2.5'], np.arange((1., 2500), 100))
-	ax.pcolormesh(_df.index, np.arange((1., 2500), 100), shade_value, cmap='binary', vmin=0, vmax=300,
-				  shading='auto')
-
-	## plot day and night
-	ld = ax.scatter(_df.index[0:], _df_day['Ext'], s=50, c='#73b9ff', label='Day Ext', marker='o', alpha=.7)
-	ln = ax.scatter(_df.index[0:], _df_night['Ext'], s=50, c='#00238c', label='Night Ext', marker='o', alpha=.7)
-
-	ax2 = ax.twinx()
-	# ld, = ax2.plot(_df_day['VC'],c='#FF9797',label='day 06:00~18:00')
-	# ln, = ax2.plot(_df_night['VC'],c='#FF0000',label='night 18:00~06:00')
-	ld2 = ax2.scatter(_df.index, _df_day['VC'], s=50, c='#FF9797', label='Day VC', marker='o', alpha=.5)
-	ln2 = ax2.scatter(_df.index, _df_night['VC'], s=50, c='#FF0000', label='Night VC', marker='o', alpha=.5)
-
-	# add legend on the first axes
-	ax.legend(handles=[ld, ln, ld2, ln2], framealpha=0, prop={'weight': 'bold'}, loc='upper left')
-
-	# add xlabel, ylabel, suptitle
-	ax.set(xlabel='Date',
-		   ylabel='Ext (1/Mm)',
-		   xlim=(st_tm, fn_tm),
-		   ylim=(1., 600),
-		   xticks=tick_time,
-		   xticklabels=[_tm.strftime("%F %H:00") for _tm in tick_time])
-
-	ax2.set(ylabel=r'$VC (m^{2}/s)$',
-			ylim=(1., 2500))
-
-	fig.suptitle(f'Event evolution ({st_tm.strftime("%F")}_{fn_tm.strftime("%F")})')
-
-	# save figure
-	fig.savefig(pth(f"event_evolution_{st_tm.strftime("%F")}_{fn_tm.strftime("%F")}"))
-
-
-if __name__ == '__main__':
-	event_evolution(blh)

AeroViz/plot/templates/regression.py

@@ -1,256 +0,0 @@
-import matplotlib.pyplot as plt
-import numpy as np
-import pandas as pd
-from matplotlib.pyplot import Figure, Axes
-from sklearn.linear_model import LinearRegression
-from tabulate import tabulate
-
-from AeroViz.plot.utils import *
-
-__all__ = [
-	'linear_regression',
-	'multiple_linear_regression',
-]
-
-
-def _linear_regression(x_array: np.ndarray,
-					   y_array: np.ndarray,
-					   columns: str | list[str] | None = None,
-					   positive: bool = True,
-					   fit_intercept: bool = True):
-	if len(x_array.shape) > 1 and x_array.shape[1] >= 2:
-		model = LinearRegression(positive=positive, fit_intercept=fit_intercept).fit(x_array, y_array)
-
-		coefficients = model.coef_[0].round(3)
-		intercept = model.intercept_[0].round(3) if fit_intercept else 'None'
-		r_square = model.score(x_array, y_array).round(3)
-		y_predict = model.predict(x_array)
-
-		equation = ' + '.join([f'{coeff:.3f} * {col}' for coeff, col in zip(coefficients, columns)])
-		equation = equation.replace(' + 0.000 * Const', '')  # Remove terms with coefficient 0
-
-		text = 'y = ' + str(equation) + '\n' + r'$\bf R^2 = $' + str(r_square)
-		tab = tabulate([[*coefficients, intercept, r_square]], headers=[*columns, 'intercept', 'R^2'], floatfmt=".3f",
-					   tablefmt="fancy_grid")
-		print('\n' + tab)
-
-		return text, y_predict, coefficients
-
-	else:
-		x_array = x_array.reshape(-1, 1)
-		y_array = y_array.reshape(-1, 1)
-
-		model = LinearRegression(positive=positive, fit_intercept=fit_intercept).fit(x_array, y_array)
-
-		slope = model.coef_[0][0].round(3)
-		intercept = model.intercept_[0].round(3) if fit_intercept else 'None'
-		r_square = model.score(x_array, y_array).round(3)
-		y_predict = model.predict(x_array)
-
-		text = np.poly1d([slope, intercept])
-		text = 'y = ' + str(text).replace('\n', "") + '\n' + r'$\bf R^2 = $' + str(r_square)
-
-		tab = tabulate([[slope, intercept, r_square]], headers=['slope', 'intercept', 'R^2'], floatfmt=".3f",
-					   tablefmt="fancy_grid")
-		print('\n' + tab)
-
-		return text, y_predict, slope
-
-
-@set_figure
-def linear_regression(df: pd.DataFrame,
-					  x: str | list[str],
-					  y: str | list[str],
-					  labels: str | list[str] = None,
-					  ax: Axes | None = None,
-					  diagonal=False,
-					  positive: bool = True,
-					  fit_intercept: bool = True,
-					  **kwargs
-					  ) -> tuple[Figure, Axes]:
-	"""
-	Create a scatter plot with multiple regression lines for the given data.
-
-	Parameters
-	----------
-	df : DataFrame
-		Input DataFrame containing the data.
-
-	x : str or list of str
-		Column name(s) for the x-axis variable(s).
-
-	y : str or list of str
-		Column name(s) for the y-axis variable(s).
-
-	labels : str or list of str, optional
-		Labels for the y-axis variable(s). If None, column names are used as labels. Default is None.
-
-	ax : AxesSubplot, optional
-		Matplotlib AxesSubplot to use for the plot. If None, a new subplot is created. Default is None.
-
-	diagonal : bool, optional
-		If True, a diagonal line (1:1 line) is added to the plot. Default is False.
-
-	positive : bool, optional
-	   Whether to let coefficient positive. Default is True.
-
-	fit_intercept: bool, optional
-		Whether to fit intercept. Default is True.
-
-	**kwargs
-		Additional keyword arguments to customize the plot.
-
-	Returns
-	-------
-	AxesSubplot
-		Matplotlib AxesSubplot containing the scatter plot.
-
-	Notes
-	-----
-	- The function creates a scatter plot with the option to include multiple regression lines.
-	- If regression is True, regression lines are fitted for each y variable.
-	- Additional customization can be done using the **kwargs.
-
-	Example
-	-------
-	>>> linear_regression(df, x='X', y=['Y1', 'Y2'], labels=['Label1', 'Label2'],
-	...                      regression=True, diagonal=True, xlim=(0, 10), ylim=(0, 20),
-	...                      xlabel="X-axis", ylabel="Y-axis", title="Scatter Plot with Regressions")
-	"""
-	fig, ax = plt.subplots(**kwargs.get('fig_kws', {})) if ax is None else (ax.get_figure(), ax)
-
-	if not isinstance(x, str):
-		x = x[0]
-
-	if not isinstance(y, list):
-		y = [y]
-
-	if labels is None:
-		labels = y
-
-	df = df.dropna(subset=[x, *y])
-	x_array = df[[x]].to_numpy()
-
-	color_cycle = Color.linecolor
-
-	handles, text_list = [], []
-
-	for i, y_var in enumerate(y):
-		y_array = df[[y_var]].to_numpy()
-
-		color = color_cycle[i % len(color_cycle)]
-
-		scatter = ax.scatter(x_array, y_array, s=25, color=color['face'], edgecolors=color['edge'], alpha=0.8,
-							 label=labels[i])
-		handles.append(scatter)
-
-		text, y_predict, slope = _linear_regression(x_array, y_array,
-													columns=labels[i],
-													positive=positive,
-													fit_intercept=fit_intercept)
-
-		text_list.append(f'{labels[i]}: {text}')
-		plt.plot(x_array, y_predict, linewidth=3, color=color['line'], alpha=1, zorder=3)
-
-	ax.set(xlim=kwargs.get('xlim'), ylim=kwargs.get('ylim'), xlabel=Unit(x), ylabel=Unit(y[0]),
-		   title=kwargs.get('title'))
-
-	# Add regression info to the legend
-	leg = plt.legend(handles=handles, labels=text_list, loc='upper left', prop={'weight': 'bold', 'size': 10})
-
-	for text, color in zip(leg.get_texts(), [color['line'] for color in color_cycle]):
-		text.set_color(color)
-
-	if diagonal:
-		ax.axline((0, 0), slope=1., color='k', lw=2, ls='--', alpha=0.5, label='1:1')
-		plt.text(0.97, 0.97, r'$\bf 1:1\ Line$', color='k', ha='right', va='top', transform=ax.transAxes)
-
-	plt.show()
-
-	return fig, ax
-
-
-@set_figure
-def multiple_linear_regression(df: pd.DataFrame,
-							   x: str | list[str],
-							   y: str | list[str],
-							   labels: str | list[str] = None,
-							   ax: Axes | None = None,
-							   diagonal=False,
-							   positive: bool = True,
-							   fit_intercept: bool = True,
-							   **kwargs
-							   ) -> tuple[Figure, Axes]:
-	"""
-	Perform multiple linear regression analysis and plot the results.
-
-	Parameters
-	----------
-	df : pandas.DataFrame
-	   Input DataFrame containing the data.
-
-	x : str or list of str
-	   Column name(s) for the independent variable(s). Can be a single string or a list of strings.
-
-	y : str or list of str
-	   Column name(s) for the dependent variable(s). Can be a single string or a list of strings.
-
-	labels : str or list of str, optional
-	   Labels for the dependent variable(s). If None, column names are used as labels. Default is None.
-
-	ax : matplotlib.axes.Axes or None, optional
-	   Matplotlib Axes object to use for the plot. If None, a new subplot is created. Default is None.
-
-	diagonal : bool, optional
-	   Whether to include a diagonal line (1:1 line) in the plot. Default is False.
-
-	positive : bool, optional
-	   Whether to let coefficient positive. Default is True.
-
-	fit_intercept: bool, optional
-		Whether to fit intercept. Default is True.
-
-	**kwargs
-	   Additional keyword arguments to customize the plot.
-
-	Returns
-	-------
-	matplotlib.axes.Axes
-	   Matplotlib Axes object containing the regression plot.
-
-	Notes
-	-----
-	This function performs multiple linear regression analysis using the input DataFrame.
-	It supports multiple independent variables and can plot the regression results.
-
-	Example
-	-------
-	>>> multiple_linear_regression(df, x=['X1', 'X2'], y='Y', labels=['Y1', 'Y2'],
-	...                             diagonal=True, add_constant=True,
-	...                             xlabel="X-axis", ylabel="Y-axis", title="Multiple Linear Regression Plot")
-	"""
-	fig, ax = plt.subplots(**kwargs.get('fig_kws', {})) if ax is None else (ax.get_figure(), ax)
-
-	if not isinstance(x, list):
-		x = [x]
-
-	if not isinstance(y, str):
-		y = y[0]
-
-	if labels is None:
-		labels = x
-
-	df = df[[*x, y]].dropna()
-	x_array = df[[*x]].to_numpy()
-	y_array = df[[y]].to_numpy()
-
-	text, y_predict, coefficients = _linear_regression(x_array, y_array,
-													   columns=labels,
-													   positive=positive,
-													   fit_intercept=fit_intercept)
-
-	df = pd.DataFrame(np.concatenate([y_array, y_predict], axis=1), columns=['y_actual', 'y_predict'])
-
-	linear_regression(df, x='y_actual', y='y_predict', ax=ax, regression=True, diagonal=diagonal)
-
-	return fig, ax