python 代码实现了一个条件生成对抗网络（Conditional Generative Adversarial Network，CGAN），用于生成与给定的理化值相关的光谱数据

import tensorflow as tf
import numpy as np
import pandas as pd
import os
import matplotlib.pyplot as plt
from sklearn.model_selection import train_test_split
from tensorflow.keras.layers import Add, BatchNormalization

os.environ["KMP_DUPLICATE_LIB_OK"] = "TRUE"

# 数据加载函数
def load_spectrum_data(csv_file):
    df = pd.read_csv(csv_file)
    wavelengths = df.columns[1:-1].astype(float)  # 确保波长是数值类型
    x = df.iloc[:, 1:-1].values  # 光谱数据
    y = df.iloc[:, -1].values

    spectra, spectra2, physicochemical_values, physicochemical_values2 = train_test_split(x,y,train_size=0.33)


    # 理化值
    return wavelengths, spectra, physicochemical_values

# 残差块函数
def residual_block(inputs, filters, kernel_size, activation):
    x = tf.keras.layers.Conv1D(filters, kernel_size, padding='same', activation=activation)(inputs)
    x = BatchNormalization()(x)
    x = tf.keras.layers.Conv1D(filters, kernel_size, padding='same', activation=None)(x)
    x = BatchNormalization()(x)
    return Add()([inputs, x])

# 构建生成器
def build_generator(z_dim, y_dim, output_dim):
    inputs_z = tf.keras.layers.Input(shape=(z_dim,))
    inputs_y = tf.keras.layers.Input(shape=(y_dim,))
    inputs = tf.keras.layers.Concatenate()([inputs_z, inputs_y])

    x = tf.keras.layers.Dense(512, activation='relu')(inputs)
    x = tf.keras.layers.Dense(256, activation='relu')(x)
    x = tf.keras.layers.Reshape((256, 1))(x)

    # 增加残差块
    x = residual_block(x, 128, kernel_size=5, activation='relu')
    x = residual_block(x, 128, kernel_size=5, activation='relu')

    x = tf.keras.layers.Conv1D(64, kernel_size=