ANNs_TRwin_functions.py

"""
*** Copyright Notice ***

PyDynamicWindow Copyright (c) 2024, The Regents of the University of California, through Lawrence Berkeley National Laboratory (subject to receipt of any required approvals from the U.S. Dept. of Energy). All rights reserved.

If you have questions about your rights to use or distribute this software,
please contact Berkeley Lab's Intellectual Property Office at
IPO@lbl.gov.

NOTICE.  This Software was developed under funding from the U.S. Department
of Energy and the U.S. Government consequently retains certain rights.  As
such, the U.S. Government has been granted for itself and others acting on
its behalf a paid-up, nonexclusive, irrevocable, worldwide license in the
Software to reproduce, distribute copies to the public, prepare derivative 
works, and perform publicly and display publicly, and to permit others to do so.
"""

# ANNs_TRwin_functions
# define 7 different functions for various inputs and outputs
# output files include npy, png, tif

import sklearn
import pandas as pd
from joblib import load
import matplotlib.pyplot as plt
import numpy as np
import rasterio
import os


# Function 01
# Output: Etr,
# Input: Tsol_dark, Tsol_clear, FileName

def world_map_Etr_TsolAsInput(Tsol_dark, Tsol_clear, FileName):
    # input solar transmittance in clear state and dark state
    target_values = np.array([Tsol_clear, Tsol_dark])
    # load scaler
    # Get the current script directory
    script_dir = os.path.dirname(__file__)
    # Construct absolute path to the file you want to load
    file_path_scaler = os.path.join(script_dir, 'ANN_joblib_files', '401_Etr_Map_2226_Random_TsolAsTrain_save_split_scaler_05.joblib')
    scaler = load(file_path_scaler)
    # load model
    file_path_model = os.path.join(script_dir, 'ANN_joblib_files',
                                    '401_Etr_Map_2226_Random_TsolAsTrain_save_split_05.joblib')

    final_model_Etr = load(file_path_model)

    # load map data
    data_path_map = 'Map_tiff_npy_files'
    filename_Tavg = 'world_input_Tavg.tif'
    filename_Tmin = 'world_input_Tmin.tif'
    filename_Tmax = 'world_input_Tmax.tif'

    file_full_Tavg = os.path.join(script_dir, data_path_map, filename_Tavg)
    file_full_Tmin = os.path.join(script_dir, data_path_map, filename_Tmin)
    file_full_Tmax = os.path.join(script_dir, data_path_map, filename_Tmax)

    # read tif files
    # Tavg
    src_Tavg = rasterio.open(file_full_Tavg)
    array_Tavg = src_Tavg.read(1)
    # Tmin
    src_Tmin = rasterio.open(file_full_Tmin)
    array_Tmin = src_Tmin.read(1)
    # Tmax
    src_Tmax = rasterio.open(file_full_Tmax)
    array_Tmax = src_Tmax.read(1)
    # lat
    height = array_Tavg.shape[0]
    width = array_Tavg.shape[1]
    cols, rows = np.meshgrid(np.arange(width), np.arange(height))
    xs, ys = rasterio.transform.xy(src_Tavg.transform, rows, cols)
    array_lons = np.array(xs)
    array_lats = np.array(ys)

    file_full_GVI = os.path.join(script_dir, data_path_map, '204_GVI_MLP_2226_grid_save.py.npy')
    array_gvi = np.load(file_full_GVI)

    list_Tavg = array_Tavg.flatten()
    list_Tmin = array_Tmin.flatten()
    list_Tmax = array_Tmax.flatten()
    list_lats = array_lats.flatten()
    list_gvi = array_gvi.flatten()
    # add Low_Tsol and High_Tsol, B = np.full_like(A, 1)
    list_Low_Tsol = np.full_like(list_Tavg, target_values[1])
    list_High_Tsol = np.full_like(list_Tavg, target_values[0])

    nan_mask = np.isnan(list_gvi) | np.isnan(list_Tavg) | np.isnan(list_Tmin) | np.isnan(list_Tmax)
    list_lats[nan_mask] = np.nan

    data_map_input = pd.DataFrame({
        'abs_latitude': np.abs(list_lats),
        'gvi': list_gvi,
        'Tair_ave': list_Tavg,
        'Tair_min': list_Tmin,
        'Tair_max': list_Tmax,
        'Low_Tsol': list_Low_Tsol,
        'High_Tsol': list_High_Tsol,
    })

    list_Etr = np.full_like(list_lats, np.nan)

    # map heatmap 
    data_map_input_scale_Etr = scaler.transform(data_map_input)
    list_Etr[~nan_mask] = final_model_Etr.predict(data_map_input_scale_Etr[~nan_mask])
    reshaped_array = list_Etr.reshape((height, width))

    # Save as a NumPy npy file
    np.save(f'{FileName}_Etr_TsolAsInput.npy', reshaped_array)
    # Draw the wide heatmap occupying the entire subplot
    fig = plt.figure(figsize=(12, 4))
    ax = fig.add_subplot(111)
    im = ax.imshow(reshaped_array, cmap='viridis', extent=[-180, 180, -55, 60], aspect='auto')
    plt.colorbar(im, ax=ax)
    ax.grid(True, linestyle='dashed', linewidth=0.5, color='lightgrey')
    ax.set_xlabel('Longitude (\u00B0)')
    ax.set_ylabel('Latitude (\u00B0)')
    ax.set_title(
        f'$\u0394E_{{TR}}$ (kWh/m$^2$) when $\u03C4_{{dark}}$ = {target_values[1]:.2f} and $\u03C4_{{clear}}$ = {target_values[0]:.2f}')
    plt.tight_layout()
    # Save the figure as a PNG file
    plt.savefig(f'{FileName}_Etr_TsolAsInput.png', dpi=300)
    # Show the figure
    # plt.show()

    # Save as a GeoTIFF file
    with rasterio.open(
            f'{FileName}_Etr_TsolAsInput.tif',
            'w',
            driver='GTiff',
            height=reshaped_array.shape[0],
            width=reshaped_array.shape[1],
            count=1,
            dtype=reshaped_array.dtype,
            crs=src_Tavg.crs,
            transform=src_Tavg.transform,
    ) as dst_Etr:
        dst_Etr.write(reshaped_array, 1)

# Function 02
# Output: En,
# Input: Tsol_dark, Tsol_clear, FileName
# for two inputs function, En should be a non-negative value because of opt Ttran

def world_map_En_TsolAsInput(Tsol_dark, Tsol_clear, FileName):
    # input solar transmittance in clear state and dark state
    target_values = np.array([Tsol_clear, Tsol_dark])
    # load scaler
    # Get the current script directory
    script_dir = os.path.dirname(__file__)
    # Construct absolute path to the file you want to load
    file_path_scaler = os.path.join(script_dir, 'ANN_joblib_files', '402_En_Map_2226_Random_TsolAsTrain_save_scaler.joblib')
    scaler = load(file_path_scaler)
    # load model
    file_path_model = os.path.join(script_dir, 'ANN_joblib_files',
                                    '402_En_Map_2226_Random_TsolAsTrain_save_02.joblib')
    final_model_En = load(file_path_model)

    # load map data
    data_path_map = 'Map_tiff_npy_files'
    filename_Tavg = 'world_input_Tavg.tif'
    filename_Tmin = 'world_input_Tmin.tif'
    filename_Tmax = 'world_input_Tmax.tif'

    file_full_Tavg = os.path.join(script_dir, data_path_map, filename_Tavg)
    file_full_Tmin = os.path.join(script_dir, data_path_map, filename_Tmin)
    file_full_Tmax = os.path.join(script_dir, data_path_map, filename_Tmax)

    # read tif files
    # Tavg
    src_Tavg = rasterio.open(file_full_Tavg)
    array_Tavg = src_Tavg.read(1)
    # Tmin
    src_Tmin = rasterio.open(file_full_Tmin)
    array_Tmin = src_Tmin.read(1)
    # Tmax
    src_Tmax = rasterio.open(file_full_Tmax)
    array_Tmax = src_Tmax.read(1)
    # lat
    height = array_Tavg.shape[0]
    width = array_Tavg.shape[1]
    cols, rows = np.meshgrid(np.arange(width), np.arange(height))
    xs, ys = rasterio.transform.xy(src_Tavg.transform, rows, cols)
    array_lons = np.array(xs)
    array_lats = np.array(ys)

    file_full_GVI = os.path.join(script_dir, data_path_map, '204_GVI_MLP_2226_grid_save.py.npy')
    array_gvi = np.load(file_full_GVI)

    list_Tavg = array_Tavg.flatten()
    list_Tmin = array_Tmin.flatten()
    list_Tmax = array_Tmax.flatten()
    list_lats = array_lats.flatten()
    list_gvi = array_gvi.flatten()
    # add Low_Tsol and High_Tsol, B = np.full_like(A, 1)
    list_Low_Tsol = np.full_like(list_Tavg, target_values[1])
    list_High_Tsol = np.full_like(list_Tavg, target_values[0])

    nan_mask = np.isnan(list_gvi) | np.isnan(list_Tavg) | np.isnan(list_Tmin) | np.isnan(list_Tmax)
    list_lats[nan_mask] = np.nan

    data_map_input = pd.DataFrame({
        'abs_latitude': np.abs(list_lats),
        'gvi': list_gvi,
        'Tair_ave': list_Tavg,
        'Tair_min': list_Tmin,
        'Tair_max': list_Tmax,
        'Low_Tsol': list_Low_Tsol,
        'High_Tsol': list_High_Tsol,
    })

    list_En = np.full_like(list_lats, np.nan)

    # map heatmap En
    data_map_input_scale_En = scaler.transform(data_map_input)
    list_En[~nan_mask] = final_model_En.predict(data_map_input_scale_En[~nan_mask])
    # non-negative value
    list_En[list_En < 0] = 0
    reshaped_array = list_En.reshape((height, width))

    # Save as a NumPy npy file
    np.save(f'{FileName}_En_TsolAsInput.npy', reshaped_array)
    # Draw the wide heatmap occupying the entire subplot
    fig = plt.figure(figsize=(12, 4))
    ax = fig.add_subplot(111)
    im = ax.imshow(reshaped_array, cmap='viridis', extent=[-180, 180, -55, 60], aspect='auto')
    plt.colorbar(im, ax=ax)
    ax.grid(True, linestyle='dashed', linewidth=0.5, color='lightgrey')
    ax.set_xlabel('Longitude (\u00B0)')
    ax.set_ylabel('Latitude (\u00B0)')
    ax.set_title(
        f'$\u0394E_{{n}}$ (kWh/m$^2$) when $\u03C4_{{dark}}$ = {target_values[1]:.2f} and $\u03C4_{{clear}}$ = {target_values[0]:.2f}')
    plt.tight_layout()
    # Save the figure as a PNG file
    plt.savefig(f'{FileName}_En_TsolAsInput.png', dpi=300)
    # Show the figure
    # plt.show()

    # Save as a GeoTIFF file
    with rasterio.open(
            f'{FileName}_En_TsolAsInput.tif',
            'w',
            driver='GTiff',
            height=reshaped_array.shape[0],
            width=reshaped_array.shape[1],
            count=1,
            dtype=reshaped_array.dtype,
            crs=src_Tavg.crs,
            transform=src_Tavg.transform,
    ) as dst_En:
        dst_En.write(reshaped_array, 1)

# Function 03
# Output: Ttran,
# Input: Tsol_dark, Tsol_clear, FileName


def world_map_Ttran_TsolAsInput(Tsol_dark, Tsol_clear, FileName):
    # input solar transmittance in clear state and dark state
    target_values = np.array([Tsol_clear, Tsol_dark])
    # load scaler
    # Get the current script directory
    script_dir = os.path.dirname(__file__)
    # Construct absolute path to the file you want to load
    file_path_scaler = os.path.join(script_dir, 'ANN_joblib_files',
                                    '403_optTtran_Map_2226_Random_TsolAsTrain_save_scaler.joblib')
    scaler = load(file_path_scaler)
    # load model
    file_path_model = os.path.join(script_dir, 'ANN_joblib_files',
                                   '403_optTtran_Map_2226_Random_TsolAsTrain_save_03.joblib')
    final_model_Ttran = load(file_path_model)

    # load map data
    data_path_map = 'Map_tiff_npy_files'
    filename_Tavg = 'world_input_Tavg.tif'
    filename_Tmin = 'world_input_Tmin.tif'
    filename_Tmax = 'world_input_Tmax.tif'

    file_full_Tavg = os.path.join(script_dir, data_path_map, filename_Tavg)
    file_full_Tmin = os.path.join(script_dir, data_path_map, filename_Tmin)
    file_full_Tmax = os.path.join(script_dir, data_path_map, filename_Tmax)

    # read tif files
    # Tavg
    src_Tavg = rasterio.open(file_full_Tavg)
    array_Tavg = src_Tavg.read(1)
    # Tmin
    src_Tmin = rasterio.open(file_full_Tmin)
    array_Tmin = src_Tmin.read(1)
    # Tmax
    src_Tmax = rasterio.open(file_full_Tmax)
    array_Tmax = src_Tmax.read(1)
    # lat
    height = array_Tavg.shape[0]
    width = array_Tavg.shape[1]
    cols, rows = np.meshgrid(np.arange(width), np.arange(height))
    xs, ys = rasterio.transform.xy(src_Tavg.transform, rows, cols)
    array_lons = np.array(xs)
    array_lats = np.array(ys)

    file_full_GVI = os.path.join(script_dir, data_path_map, '204_GVI_MLP_2226_grid_save.py.npy')
    array_gvi = np.load(file_full_GVI)

    list_Tavg = array_Tavg.flatten()
    list_Tmin = array_Tmin.flatten()
    list_Tmax = array_Tmax.flatten()
    list_lats = array_lats.flatten()
    list_gvi = array_gvi.flatten()
    # add Low_Tsol and High_Tsol, B = np.full_like(A, 1)
    list_Low_Tsol = np.full_like(list_Tavg, target_values[1])
    list_High_Tsol = np.full_like(list_Tavg, target_values[0])

    nan_mask = np.isnan(list_gvi) | np.isnan(list_Tavg) | np.isnan(list_Tmin) | np.isnan(list_Tmax)
    list_lats[nan_mask] = np.nan

    data_map_input = pd.DataFrame({
        'abs_latitude': np.abs(list_lats),
        'gvi': list_gvi,
        'Tair_ave': list_Tavg,
        'Tair_min': list_Tmin,
        'Tair_max': list_Tmax,
        'Low_Tsol': list_Low_Tsol,
        'High_Tsol': list_High_Tsol,
    })

    list_Ttran = np.full_like(list_lats, np.nan)

    # map heatmap En
    data_map_input_scale_Ttran = scaler.transform(data_map_input)
    list_Ttran[~nan_mask] = final_model_Ttran.predict(data_map_input_scale_Ttran[~nan_mask])
    reshaped_array = list_Ttran.reshape((height, width))

    # Save as a NumPy npy file
    np.save(f'{FileName}_Ttran_TsolAsInput.npy', reshaped_array)
    # Draw the wide heatmap occupying the entire subplot
    fig = plt.figure(figsize=(12, 4))
    ax = fig.add_subplot(111)
    im = ax.imshow(reshaped_array, cmap='viridis', extent=[-180, 180, -55, 60], aspect='auto')
    plt.colorbar(im, ax=ax)
    ax.grid(True, linestyle='dashed', linewidth=0.5, color='lightgrey')
    ax.set_xlabel('Longitude (\u00B0)')
    ax.set_ylabel('Latitude (\u00B0)')
    ax.set_title(
        f'Opt $T_{{tran}}$ (\u00B0C) when $\u03C4_{{dark}}$ = {target_values[1]:.2f} and $\u03C4_{{clear}}$ = {target_values[0]:.2f}')
    plt.tight_layout()
    # Save the figure as a PNG file
    plt.savefig(f'{FileName}_Ttran_TsolAsInput.png', dpi=300)
    # Show the figure
    # plt.show()

    # Save as a GeoTIFF file
    with rasterio.open(
            f'{FileName}_Ttran_TsolAsInput.tif',
            'w',
            driver='GTiff',
            height=reshaped_array.shape[0],
            width=reshaped_array.shape[1],
            count=1,
            dtype=reshaped_array.dtype,
            crs=src_Tavg.crs,
            transform=src_Tavg.transform,
    ) as dst_Ttran:
        dst_Ttran.write(reshaped_array, 1)

# Function 04
# Output: TRRI,
# Input: Tsol_dark, Tsol_clear, Etr_npy_array, En_npy_array, FileName

def world_map_TRRI_TsolAsInput(Tsol_dark, Tsol_clear, Etr_npy_array, En_npy_array, FileName):
    # input solar transmittance in clear state and dark state
    target_values = np.array([Tsol_clear, Tsol_dark])
    # Get the current script directory
    script_dir = os.path.dirname(__file__)
    # load Etr and En
    array_Etr = np.load(Etr_npy_array)
    array_En = np.load(En_npy_array)
    # array to list
    list_Etr = array_Etr.flatten()
    list_En = array_En.flatten()
    # ReLU
    list_Etr[list_Etr < 0] = 0
    list_En[list_En < 0] = 0

    # load Tavg
    data_path_map = 'Map_tiff_npy_files'
    filename_Tavg = 'world_input_Tavg.tif'
    file_full_Tavg = os.path.join(script_dir, data_path_map, filename_Tavg)
    # read tif files
    # Tavg
    src_Tavg = rasterio.open(file_full_Tavg)

    # define empty list
    list_TRRI = list()
    # loop
    for value_Etr, value_En in zip(list_Etr, list_En):
        if (np.isnan(value_Etr) or np.isnan(value_En)):
            list_TRRI.append(np.nan)
        else:
            temp_input = value_Etr*value_En
            list_TRRI.append(temp_input)

    # reshape list
    reshaped_array = np.reshape(list_TRRI, (690, 2160))

    # Save as a NumPy npy file
    np.save(f'{FileName}_TRRI_TsolAsInput.npy', reshaped_array)
    # Draw the wide heatmap occupying the entire subplot
    fig = plt.figure(figsize=(12, 4))
    ax = fig.add_subplot(111)
    im = ax.imshow(reshaped_array, cmap='viridis', extent=[-180, 180, -55, 60], aspect='auto')
    plt.colorbar(im, ax=ax)
    ax.grid(True, linestyle='dashed', linewidth=0.5, color='lightgrey')
    ax.set_xlabel('Longitude (\u00B0)')
    ax.set_ylabel('Latitude (\u00B0)')
    ax.set_title(
        f'TRRI (kWh$^2$/m$^4$) when $\u03C4_{{dark}}$ = {target_values[1]:.2f} and $\u03C4_{{clear}}$ = {target_values[0]:.2f}')
    plt.tight_layout()
    # Save the figure as a PNG file
    plt.savefig(f'{FileName}_TRRI_TsolAsInput.png', dpi=300)
    # Show the figure
    # plt.show()

    # Save as a GeoTIFF file
    with rasterio.open(
            f'{FileName}_TRRI_TsolAsInput.tif',
            'w',
            driver='GTiff',
            height=reshaped_array.shape[0],
            width=reshaped_array.shape[1],
            count=1,
            dtype=reshaped_array.dtype,
            crs=src_Tavg.crs,
            transform=src_Tavg.transform,
    ) as dst_TRRI:
        dst_TRRI.write(reshaped_array, 1)

# Function 05
# Output: Etr,
# Input: Tsol_dark, Tsol_clear, Ttran, FileName

def world_map_Etr_TsolTtranAsInput(Tsol_dark, Tsol_clear, Ttran, FileName):
    # input solar transmittance in clear state and dark state, and transition temperature
    target_values = np.array([Tsol_clear, Tsol_dark, Ttran])
    # load scaler
    # Get the current script directory
    script_dir = os.path.dirname(__file__)
    # Construct absolute path to the file you want to load
    file_path_scaler = os.path.join(script_dir, 'ANN_joblib_files',
                                    '501_Etr_Map_2226_Random_AlllAsTrain_save_scaler.joblib')
    scaler = load(file_path_scaler)
    # load model
    file_path_model = os.path.join(script_dir, 'ANN_joblib_files',
                                   '501_Etr_Map_2226_Random_AlllAsTrain_save_01.joblib')
    final_model_Etr = load(file_path_model)

    # load map data
    data_path_map = 'Map_tiff_npy_files'
    filename_Tavg = 'world_input_Tavg.tif'
    filename_Tmin = 'world_input_Tmin.tif'
    filename_Tmax = 'world_input_Tmax.tif'

    file_full_Tavg = os.path.join(script_dir, data_path_map, filename_Tavg)
    file_full_Tmin = os.path.join(script_dir, data_path_map, filename_Tmin)
    file_full_Tmax = os.path.join(script_dir, data_path_map, filename_Tmax)

    # read tif files
    # Tavg
    src_Tavg = rasterio.open(file_full_Tavg)
    array_Tavg = src_Tavg.read(1)
    # Tmin
    src_Tmin = rasterio.open(file_full_Tmin)
    array_Tmin = src_Tmin.read(1)
    # Tmax
    src_Tmax = rasterio.open(file_full_Tmax)
    array_Tmax = src_Tmax.read(1)
    # lat
    height = array_Tavg.shape[0]
    width = array_Tavg.shape[1]
    cols, rows = np.meshgrid(np.arange(width), np.arange(height))
    xs, ys = rasterio.transform.xy(src_Tavg.transform, rows, cols)
    array_lons = np.array(xs)
    array_lats = np.array(ys)

    file_full_GVI = os.path.join(script_dir, data_path_map, '204_GVI_MLP_2226_grid_save.py.npy')
    array_gvi = np.load(file_full_GVI)

    list_Tavg = array_Tavg.flatten()
    list_Tmin = array_Tmin.flatten()
    list_Tmax = array_Tmax.flatten()
    list_lats = array_lats.flatten()
    list_gvi = array_gvi.flatten()
    # add Low_Tsol and High_Tsol, B = np.full_like(A, 1)
    list_Low_Tsol = np.full_like(list_Tavg, target_values[1])
    list_High_Tsol = np.full_like(list_Tavg, target_values[0])
    list_Ttran = np.full_like(list_Tavg, target_values[2])

    nan_mask = np.isnan(list_gvi) | np.isnan(list_Tavg) | np.isnan(list_Tmin) | np.isnan(list_Tmax)
    list_lats[nan_mask] = np.nan

    data_map_input = pd.DataFrame({
        'abs_latitude': np.abs(list_lats),
        'gvi': list_gvi,
        'Tair_ave': list_Tavg,
        'Tair_min': list_Tmin,
        'Tair_max': list_Tmax,
        'Low_Tsol': list_Low_Tsol,
        'High_Tsol': list_High_Tsol,
        'TransitionTemperature': list_Ttran,
    })

    list_Etr = np.full_like(list_lats, np.nan)

    # map heatmap Etr
    data_map_input_scale_Etr = scaler.transform(data_map_input)
    list_Etr[~nan_mask] = final_model_Etr.predict(data_map_input_scale_Etr[~nan_mask])
    reshaped_array = list_Etr.reshape((height, width))

    # Save as a NumPy npy file
    np.save(f'{FileName}_Etr_TsolTtranAsInput.npy', reshaped_array)
    # Draw the wide heatmap occupying the entire subplot
    fig = plt.figure(figsize=(12, 4))
    ax = fig.add_subplot(111)
    im = ax.imshow(reshaped_array, cmap='viridis', extent=[-180, 180, -55, 60], aspect='auto')
    plt.colorbar(im, ax=ax)
    ax.grid(True, linestyle='dashed', linewidth=0.5, color='lightgrey')
    ax.set_xlabel('Longitude (\u00B0)')
    ax.set_ylabel('Latitude (\u00B0)')
    ax.set_title(
        f'$\u0394E_{{TR}}$ (kWh/m$^2$) when $\u03C4_{{dark}}$ = {target_values[1]:.2f}, $\u03C4_{{clear}}$ = {target_values[0]:.2f}, and $T_{{tran}}$ = {target_values[2]:.2f}\u00B0C')
    plt.tight_layout()
    # Save the figure as a PNG file
    plt.savefig(f'{FileName}_Etr_TsolTtranAsInput.png', dpi=300)
    # Show the figure
    # plt.show()

    # Save as a GeoTIFF file
    with rasterio.open(
            f'{FileName}_Etr_TsolTtranAsInput.tif',
            'w',
            driver='GTiff',
            height=reshaped_array.shape[0],
            width=reshaped_array.shape[1],
            count=1,
            dtype=reshaped_array.dtype,
            crs=src_Tavg.crs,
            transform=src_Tavg.transform,
    ) as dst_Etr:
        dst_Etr.write(reshaped_array, 1)

# Function 06
# Output: En,
# Input: Tsol_dark, Tsol_clear, Ttran, FileName

def world_map_En_TsolTtranAsInput(Tsol_dark, Tsol_clear, Ttran, FileName):
    # input solar transmittance in clear state and dark state, and transition temperature
    target_values = np.array([Tsol_clear, Tsol_dark, Ttran])
    # load scaler
    # Get the current script directory
    script_dir = os.path.dirname(__file__)
    # Construct absolute path to the file you want to load
    file_path_scaler = os.path.join(script_dir, 'ANN_joblib_files',
                                    '502_En_Map_2226_Random_AlllAsTrain_save_scaler.joblib')
    scaler = load(file_path_scaler)
    # load model
    file_path_model = os.path.join(script_dir, 'ANN_joblib_files',
                                   '502_En_Map_2226_Random_AlllAsTrain_save_01.joblib')
    final_model_En = load(file_path_model)

    # load map data
    data_path_map = 'Map_tiff_npy_files'
    filename_Tavg = 'world_input_Tavg.tif'
    filename_Tmin = 'world_input_Tmin.tif'
    filename_Tmax = 'world_input_Tmax.tif'

    file_full_Tavg = os.path.join(script_dir, data_path_map, filename_Tavg)
    file_full_Tmin = os.path.join(script_dir, data_path_map, filename_Tmin)
    file_full_Tmax = os.path.join(script_dir, data_path_map, filename_Tmax)

    # read tif files
    # Tavg
    src_Tavg = rasterio.open(file_full_Tavg)
    array_Tavg = src_Tavg.read(1)
    # Tmin
    src_Tmin = rasterio.open(file_full_Tmin)
    array_Tmin = src_Tmin.read(1)
    # Tmax
    src_Tmax = rasterio.open(file_full_Tmax)
    array_Tmax = src_Tmax.read(1)
    # lat
    height = array_Tavg.shape[0]
    width = array_Tavg.shape[1]
    cols, rows = np.meshgrid(np.arange(width), np.arange(height))
    xs, ys = rasterio.transform.xy(src_Tavg.transform, rows, cols)
    array_lons = np.array(xs)
    array_lats = np.array(ys)

    file_full_GVI = os.path.join(script_dir, data_path_map, '204_GVI_MLP_2226_grid_save.py.npy')
    array_gvi = np.load(file_full_GVI)

    list_Tavg = array_Tavg.flatten()
    list_Tmin = array_Tmin.flatten()
    list_Tmax = array_Tmax.flatten()
    list_lats = array_lats.flatten()
    list_gvi = array_gvi.flatten()
    # add Low_Tsol and High_Tsol, B = np.full_like(A, 1)
    list_Low_Tsol = np.full_like(list_Tavg, target_values[1])
    list_High_Tsol = np.full_like(list_Tavg, target_values[0])
    list_Ttran = np.full_like(list_Tavg, target_values[2])

    nan_mask = np.isnan(list_gvi) | np.isnan(list_Tavg) | np.isnan(list_Tmin) | np.isnan(list_Tmax)
    list_lats[nan_mask] = np.nan

    data_map_input = pd.DataFrame({
        'abs_latitude': np.abs(list_lats),
        'gvi': list_gvi,
        'Tair_ave': list_Tavg,
        'Tair_min': list_Tmin,
        'Tair_max': list_Tmax,
        'Low_Tsol': list_Low_Tsol,
        'High_Tsol': list_High_Tsol,
        'TransitionTemperature': list_Ttran,
    })

    list_En = np.full_like(list_lats, np.nan)

    # map heatmap En
    data_map_input_scale_En = scaler.transform(data_map_input)
    list_En[~nan_mask] = final_model_En.predict(data_map_input_scale_En[~nan_mask])
    reshaped_array = list_En.reshape((height, width))

    # Save as a NumPy npy file
    np.save(f'{FileName}_En_TsolTtranAsInput.npy', reshaped_array)
    # Draw the wide heatmap occupying the entire subplot
    fig = plt.figure(figsize=(12, 4))
    ax = fig.add_subplot(111)
    im = ax.imshow(reshaped_array, cmap='viridis', extent=[-180, 180, -55, 60], aspect='auto')
    plt.colorbar(im, ax=ax)
    ax.grid(True, linestyle='dashed', linewidth=0.5, color='lightgrey')
    ax.set_xlabel('Longitude (\u00B0)')
    ax.set_ylabel('Latitude (\u00B0)')
    ax.set_title(
        f'$\u0394E_{{n}}$ (kWh/m$^2$) when $\u03C4_{{dark}}$ = {target_values[1]:.2f}, $\u03C4_{{clear}}$ = {target_values[0]:.2f}, and $T_{{tran}}$ = {target_values[2]:.2f}\u00B0C')
    plt.tight_layout()
    # Save the figure as a PNG file
    plt.savefig(f'{FileName}_En_TsolTtranAsInput.png', dpi=300)
    # Show the figure
    # plt.show()

    # Save as a GeoTIFF file
    with rasterio.open(
            f'{FileName}_En_TsolTtranAsInput.tif',
            'w',
            driver='GTiff',
            height=reshaped_array.shape[0],
            width=reshaped_array.shape[1],
            count=1,
            dtype=reshaped_array.dtype,
            crs=src_Tavg.crs,
            transform=src_Tavg.transform,
    ) as dst_En:
        dst_En.write(reshaped_array, 1)

# Function 07
# Output: TRRI,
# Input: Tsol_dark, Tsol_clear, Etr_npy_array, En_npy_array, FileName

def world_map_TRRI_TsolTtranAsInput(Tsol_dark, Tsol_clear, Ttran, Etr_npy_array, En_npy_array, FileName):
    # input solar transmittance in clear state and dark state
    target_values = np.array([Tsol_clear, Tsol_dark, Ttran])
    # Get the current script directory
    script_dir = os.path.dirname(__file__)
    # load Etr and En
    array_Etr = np.load(Etr_npy_array)
    array_En = np.load(En_npy_array)
    # array to list
    list_Etr = array_Etr.flatten()
    list_En = array_En.flatten()
    # ReLU
    list_Etr[list_Etr < 0] = 0
    list_En[list_En < 0] = 0

    # load Tavg
    data_path_map = 'Map_tiff_npy_files'
    filename_Tavg = 'world_input_Tavg.tif'
    file_full_Tavg = os.path.join(script_dir, data_path_map, filename_Tavg)
    # read tif files
    # Tavg
    src_Tavg = rasterio.open(file_full_Tavg)

    # define empty list
    list_TRRI = list()
    # loop
    for value_Etr, value_En in zip(list_Etr, list_En):
        if (np.isnan(value_Etr) or np.isnan(value_En)):
            list_TRRI.append(np.nan)
        else:
            temp_input = value_Etr*value_En
            list_TRRI.append(temp_input)

    # reshape list
    reshaped_array = np.reshape(list_TRRI, (690, 2160))

    # Save as a NumPy npy file
    np.save(f'{FileName}_TRRI_TsolAsInput.npy', reshaped_array)
    # Draw the wide heatmap occupying the entire subplot
    fig = plt.figure(figsize=(12, 4))
    ax = fig.add_subplot(111)
    im = ax.imshow(reshaped_array, cmap='viridis', extent=[-180, 180, -55, 60], aspect='auto')
    plt.colorbar(im, ax=ax)
    ax.grid(True, linestyle='dashed', linewidth=0.5, color='lightgrey')
    ax.set_xlabel('Longitude (\u00B0)')
    ax.set_ylabel('Latitude (\u00B0)')
    ax.set_title(
        f'TRRI (kWh$^2$/m$^4$) when $\u03C4_{{dark}}$ = {target_values[1]:.2f}, $\u03C4_{{clear}}$ = {target_values[0]:.2f}, and $T_{{tran}}$ = {target_values[2]:.2f}\u00B0C')
    plt.tight_layout()
    # Save the figure as a PNG file
    plt.savefig(f'{FileName}_TRRI_TsolAsInput.png', dpi=300)
    # Show the figure
    # plt.show()

    # Save as a GeoTIFF file
    with rasterio.open(
            f'{FileName}_TRRI_TsolAsInput.tif',
            'w',
            driver='GTiff',
            height=reshaped_array.shape[0],
            width=reshaped_array.shape[1],
            count=1,
            dtype=reshaped_array.dtype,
            crs=src_Tavg.crs,
            transform=src_Tavg.transform,
    ) as dst_TRRI:
        dst_TRRI.write(reshaped_array, 1)