|
|
import numpy as np
|
|
|
import json, os, shutil, bisect, subprocess, sys, time
|
|
|
from geopy.distance import geodesic
|
|
|
from netCDF4 import Dataset
|
|
|
from osgeo import gdal
|
|
|
|
|
|
# 获取当前文件的父目录的绝对路径
|
|
|
current_dir = os.path.dirname(os.path.abspath(__file__))
|
|
|
# 获取父目录的绝对路径
|
|
|
parent_dir = os.path.abspath(os.path.join(current_dir, os.pardir))
|
|
|
# 将父目录添加到 Python 模块搜索路径中
|
|
|
sys.path.append(parent_dir)
|
|
|
|
|
|
from Utils import utils as globalUtils
|
|
|
|
|
|
|
|
|
# 求2个数组的交集
|
|
|
def intersection(list1, list2):
|
|
|
set1 = set(list1)
|
|
|
set2 = set(list2)
|
|
|
return list(set1 & set2)
|
|
|
|
|
|
|
|
|
# 读取一个文件的内容，文件的格式是 “10,20,30.....”
|
|
|
def readValueListFromTxt(filePath):
|
|
|
valueList = []
|
|
|
with open(filePath, "r") as file:
|
|
|
for line in file:
|
|
|
value = [float(x) for x in line.split(",")]
|
|
|
valueList += value
|
|
|
return valueList
|
|
|
|
|
|
|
|
|
# 测试矩阵
|
|
|
def matriTest():
|
|
|
|
|
|
# 两个示例数组
|
|
|
array1 = [1, 2, 3]
|
|
|
array2 = [4, 5, 6]
|
|
|
|
|
|
# 将数组转换为NumPy数组
|
|
|
array1_np = np.array(array1)
|
|
|
array2_np = np.array(array2)
|
|
|
|
|
|
# 创建矩阵，其中每一行分别是数组1和数组2
|
|
|
matrix = np.vstack((array1_np, array2_np))
|
|
|
|
|
|
print("Matrix:")
|
|
|
print(matrix)
|
|
|
|
|
|
pass
|
|
|
|
|
|
|
|
|
# 计算一个数组中与目标值差值绝对值最小的值
|
|
|
def calClosestValue(target, arrList: list):
|
|
|
# 定义数组和目标值
|
|
|
arr = np.array(arrList)
|
|
|
|
|
|
# 计算数组中每个元素与目标值的绝对差值
|
|
|
differences = np.abs(arr - target)
|
|
|
|
|
|
# 找到差值最小的元素的索引
|
|
|
minDifferenceIndex = np.argmin(differences)
|
|
|
|
|
|
# 获取差值最小的元素
|
|
|
closestValue = arr[minDifferenceIndex]
|
|
|
|
|
|
return closestValue
|
|
|
|
|
|
|
|
|
# 找到最近的值
|
|
|
def searchNearest(sourceList: list, computeList: list) -> list:
|
|
|
closestValueList = []
|
|
|
for v in sourceList:
|
|
|
closestValueList.append(calClosestValue(v, computeList))
|
|
|
return closestValueList
|
|
|
|
|
|
|
|
|
# 计算2个点的地理距离
|
|
|
def calDistance(pointA, pointB) -> float:
|
|
|
return geodesic(pointA, pointB).miles # 使用geodesic计算距离
|
|
|
|
|
|
|
|
|
# 读取nc文件内容
|
|
|
def readNC(ncFilePath: str):
|
|
|
|
|
|
print(os.getcwd())
|
|
|
print(ncFilePath)
|
|
|
# 打开 .nc 文件，'r 表示只读
|
|
|
ncData = Dataset(ncFilePath, "r")
|
|
|
# 查看文件的维度
|
|
|
print("Dimensions:", ncData.dimensions)
|
|
|
|
|
|
# 查看文件的变量
|
|
|
print("Variables:", list(ncData.variables.keys()))
|
|
|
|
|
|
# # 获取所有的变量名称
|
|
|
# variables = ncData.variables.keys()
|
|
|
|
|
|
# # 根据变量名称获取数据
|
|
|
longitude = ncData.variables["longitude"][:]
|
|
|
latitude = ncData.variables["latitude"][:]
|
|
|
print("longitude ", len(longitude))
|
|
|
print("latitude ", len(latitude))
|
|
|
print("tco3 ", len(ncData.variables["tco3"][:]))
|
|
|
print("tcw ", len(ncData.variables["tcw"][:]))
|
|
|
|
|
|
tco3Matrix = ncData.variables["tco3"][0]
|
|
|
# 是 721 个元素的数组
|
|
|
print(len(tco3Matrix))
|
|
|
# 每个元素是 1440 长度的数组
|
|
|
print(len(tco3Matrix[0]))
|
|
|
|
|
|
tcwMatrix = ncData.variables["tcw"][0]
|
|
|
# 是 721 个元素的数组
|
|
|
print(len(tcwMatrix[0]))
|
|
|
# 每个元素是 1440 长度的数组
|
|
|
print(len(tcwMatrix[0]))
|
|
|
|
|
|
# 获取特定变量的值
|
|
|
# variable_name = 'temperature'
|
|
|
# variable_values = nc_file.variables[variable_name][:] # 这里假设 'temperature' 是文件中的一个变量名
|
|
|
|
|
|
# 关闭 .nc 文件
|
|
|
ncData.close()
|
|
|
|
|
|
return list(longitude), list(latitude), tcwMatrix, tco3Matrix
|
|
|
|
|
|
|
|
|
# 读取tif文件内容
|
|
|
def readTif(tifFilePath: str):
|
|
|
dataset = gdal.Open(tifFilePath)
|
|
|
band = dataset.GetRasterBand(1)
|
|
|
|
|
|
if dataset is not None:
|
|
|
# 获取tif文件的投影和变换信息，用于将像素位置转化为地理坐标
|
|
|
print("Projection: ", dataset.GetProjection())
|
|
|
geotransform = dataset.GetGeoTransform()
|
|
|
print("GeoTransform: ", geotransform)
|
|
|
|
|
|
# 对第一波段的数据进行读取（这里假定大气数据在第一波段）
|
|
|
band = dataset.GetRasterBand(1)
|
|
|
aodMatrix = band.ReadAsArray()
|
|
|
nrows, ncols = aodMatrix.shape
|
|
|
|
|
|
# 计算每一个像素对应的经纬度
|
|
|
longitude_left_upper = geotransform[0]
|
|
|
latitude_left_upper = geotransform[3]
|
|
|
# 纬度是-90 到 90
|
|
|
if latitude_left_upper > 90:
|
|
|
latitude_left_upper = 90
|
|
|
|
|
|
x_resolution = geotransform[1] # x像素分辨率
|
|
|
y_resolution = geotransform[5] # y像素分辨率（这个值通常为负）
|
|
|
|
|
|
# 根据左上角点坐标以及分辨率计算其他点坐标
|
|
|
lon_list = [longitude_left_upper + i * x_resolution for i in range(ncols)]
|
|
|
lat_list = [latitude_left_upper + i * y_resolution for i in range(nrows)]
|
|
|
|
|
|
print("lon_list", len(lon_list))
|
|
|
print("lat_list", len(lat_list))
|
|
|
|
|
|
# 在此可以根据计算得到的经纬度信息对大气数据进行处理
|
|
|
# 例如简单地输出数据的形状和基本统计信息
|
|
|
# print('Array shape: ', aodMatrix.shape)
|
|
|
# print('Min value: ', array.min())
|
|
|
# print('Max value: ', array.max())
|
|
|
# print('Mean value: ', array.mean())
|
|
|
|
|
|
return lon_list, lat_list, aodMatrix
|
|
|
else:
|
|
|
print("Could not open the tif file!")
|
|
|
|
|
|
pass
|
|
|
|
|
|
|
|
|
# 读取地表文件中的经纬度
|
|
|
def readSurfaceCoordinate(
|
|
|
latFilePath: str,
|
|
|
lonFilePath: str,
|
|
|
saFilePath: str,
|
|
|
szFilePath: str,
|
|
|
vaFilePath: str,
|
|
|
vzFilePath: str,
|
|
|
):
|
|
|
latList = readValueListFromTxt(latFilePath)
|
|
|
lonList = readValueListFromTxt(lonFilePath)
|
|
|
szList = readValueListFromTxt(szFilePath)
|
|
|
vzList = readValueListFromTxt(vzFilePath)
|
|
|
saList = readValueListFromTxt(saFilePath)
|
|
|
vaList = readValueListFromTxt(vaFilePath)
|
|
|
|
|
|
return lonList, latList, szList, vzList, saList, vaList
|
|
|
|
|
|
|
|
|
# 初始化输入参数文件
|
|
|
def initInputFiles(projectSurfaceDir: str, projectTOADir: str) -> bool:
|
|
|
# 地表的经纬度文件
|
|
|
surFileList = [
|
|
|
"input_surface.txt",
|
|
|
"Lat.txt",
|
|
|
"Lon.txt",
|
|
|
"SA.txt",
|
|
|
"SZ.txt",
|
|
|
"VA.txt",
|
|
|
"VZ.txt",
|
|
|
]
|
|
|
surFilePathList = [os.path.join(projectSurfaceDir, x) for x in surFileList]
|
|
|
|
|
|
# 大气的参数、nc和tif文件
|
|
|
toaFileList = ["input_toa.txt", "input.nc", "input.tif"]
|
|
|
toaFilePathList = [os.path.join(projectTOADir, x) for x in toaFileList]
|
|
|
|
|
|
# 判断这些文件是否存在，如果有1个不存在进行报错
|
|
|
allInputFileList = surFilePathList + toaFilePathList
|
|
|
for inputFile in allInputFileList:
|
|
|
if not os.path.exists(inputFile):
|
|
|
print(f"{inputFile} is not exist!")
|
|
|
return False
|
|
|
|
|
|
# 本地input目录
|
|
|
inputDir = "input"
|
|
|
if not globalUtils.rmDir(inputDir):
|
|
|
return False
|
|
|
if not globalUtils.mkDir(inputDir):
|
|
|
return False
|
|
|
|
|
|
# 拷贝参数文件到目录下
|
|
|
for inputFile in allInputFileList:
|
|
|
if not globalUtils.copyFile(
|
|
|
inputFile, os.path.join(inputDir, os.path.basename(inputFile))
|
|
|
):
|
|
|
return False
|
|
|
|
|
|
return True
|
|
|
|
|
|
|
|
|
# 获取target在arr中的最接近左值和右值，arr是正序排序
|
|
|
def getLRValue(target, arr: list):
|
|
|
# 使用bisect_left找到该数值应当插入的位置以维持有序
|
|
|
index = bisect.bisect_left(arr, target)
|
|
|
|
|
|
# 左邻近值
|
|
|
leftValue = arr[index - 1] if index > 0 else None
|
|
|
|
|
|
# 右邻近值
|
|
|
rightValue = arr[index] if index < len(arr) else None
|
|
|
|
|
|
return leftValue, rightValue
|
|
|
|
|
|
|
|
|
# 获取一个点在网格中的最接近点（网格的顶点）
|
|
|
# 注意左值和右值可能是None
|
|
|
def getNearestPoint(targetPoint, lonList, latList):
|
|
|
lonList = sorted([x for x in lonList if x is not None])
|
|
|
latList = sorted([x for x in latList if x is not None])
|
|
|
|
|
|
pointList = [(lat, lon) for lat in latList for lon in lonList]
|
|
|
# print(pointList)
|
|
|
if len(pointList) == 0:
|
|
|
return None, None
|
|
|
|
|
|
minPoint = None
|
|
|
minDistance = None
|
|
|
for point in pointList:
|
|
|
# print(targetPoint, point)
|
|
|
distance = calDistance(targetPoint, point)
|
|
|
if not minDistance or distance < minDistance:
|
|
|
minDistance = distance
|
|
|
minPoint = point
|
|
|
return minPoint, minDistance
|
|
|
|
|
|
|
|
|
def readInputSurface(inputSurfaceFilePath: str):
|
|
|
# 打开文件
|
|
|
with open(inputSurfaceFilePath, "r") as file:
|
|
|
# 使用readlines()方法读取所有行，并将结果存储在lines列表中
|
|
|
lines = file.readlines()
|
|
|
# 检查文件是否至少有两行
|
|
|
if len(lines) >= 2:
|
|
|
# 选择第二行（索引为1）
|
|
|
secondLine = lines[1]
|
|
|
splitList = secondLine.split(",")
|
|
|
if len(splitList) == 3:
|
|
|
month, wavestart, waveend = (int(x) for x in tuple(splitList))
|
|
|
# print("第二行内容:", secondLine)
|
|
|
# print(month, wavestart, waveend)
|
|
|
return month, wavestart, waveend
|
|
|
|
|
|
return None, None, None
|
|
|
# return month, wavestart, waveend
|
|
|
|
|
|
|
|
|
def readInputTOA(inputTOAFilePath: str):
|
|
|
# return aodmodel, atmodel
|
|
|
# 打开文件
|
|
|
with open(inputTOAFilePath, "r") as file:
|
|
|
# 使用readlines()方法读取所有行，并将结果存储在lines列表中
|
|
|
lines = file.readlines()
|
|
|
# 检查文件是否至少有两行
|
|
|
if len(lines) >= 2:
|
|
|
# 选择第二行（索引为1）
|
|
|
secondLine = lines[1]
|
|
|
splitList = [x for x in secondLine.split(" ") if x.strip()]
|
|
|
print(splitList)
|
|
|
if len(splitList) == 2:
|
|
|
aodmodel, atmmodel = (int(x) for x in tuple(splitList))
|
|
|
# print("第二行内容:", secondLine)
|
|
|
# print(month, wavestart, waveend)
|
|
|
return aodmodel, atmmodel
|
|
|
|
|
|
return None, None
|
|
|
|
|
|
|
|
|
# 大气每次运行需要一个地表参数文件
|
|
|
def writeSurfaceInputFile(
|
|
|
filePath: str,
|
|
|
month: int,
|
|
|
lon: float,
|
|
|
lat: float,
|
|
|
sz: float,
|
|
|
vz: float,
|
|
|
sa: float,
|
|
|
va: float,
|
|
|
wavestart: int,
|
|
|
waveend: int,
|
|
|
):
|
|
|
with open(filePath, "w") as file:
|
|
|
file.write("month,lon,lat,sz,vz,sa,va,wavestart,waveend\n")
|
|
|
file.write(f"{month},{lon},{lat},{sz},{vz},{sa},{va},{wavestart},{waveend}")
|
|
|
return filePath
|
|
|
|
|
|
|
|
|
# 大气每次运行需要一个地表参数文件
|
|
|
def writeTOAInputFile(
|
|
|
filePath: str,
|
|
|
sz: float,
|
|
|
vz: float,
|
|
|
sa: float,
|
|
|
va: float,
|
|
|
wavestart: int,
|
|
|
waveend: int,
|
|
|
aodmodel: int,
|
|
|
atmmodel: int,
|
|
|
aod: float,
|
|
|
wv: float,
|
|
|
ozone: float,
|
|
|
inputpath: str,
|
|
|
):
|
|
|
with open(filePath, "w") as file:
|
|
|
file.write(
|
|
|
"sz vz sa va wavestart waveend aodmodel atmmodel aod wv ozone inputpath\n"
|
|
|
)
|
|
|
file.write(
|
|
|
f"{sz} {vz} {sa} {va} {wavestart} {waveend} {aodmodel} {atmmodel} {aod} {wv/10} {ozone} {inputpath}"
|
|
|
)
|
|
|
return filePath
|
|
|
|
|
|
|
|
|
# 执行一次求解
|
|
|
def executeCompute():
|
|
|
# 先删除旧的output.txt
|
|
|
globalUtils.rmPath("TOASimu.txt")
|
|
|
# 启动求解任务
|
|
|
# result = subprocess.run(['AllSensorSimu.exe'])
|
|
|
process = subprocess.Popen(["AllSensorSimu.exe"])
|
|
|
|
|
|
# 等待2秒后关闭
|
|
|
while 1:
|
|
|
time.sleep(10)
|
|
|
if globalUtils.isFileOpen("TOASimu.txt"):
|
|
|
# 正在被其它程序写入，所以等待
|
|
|
print("writting TOASimu.txt")
|
|
|
else:
|
|
|
if process.poll() is None: # 如果进程还在运行，就终止它
|
|
|
process.kill()
|
|
|
print("write TOASimu.txt over!")
|
|
|
break
|
|
|
|
|
|
|
|
|
# 将一个数组归一化，非排序的数组，每个值都缩小到 [0,1]
|
|
|
def normalized(valueList: list):
|
|
|
arr = np.array(valueList)
|
|
|
return (arr - np.min(arr)) / (np.max(arr) - np.min(arr))
|
|
|
|
|
|
|
|
|
# 每个元素包含3个值，分别对应rgb三个颜色
|
|
|
# valueList: [(value445: float, value565: float, value670: float)]
|
|
|
def normalizedColor(valueList: list):
|
|
|
# 第1步，归一化
|
|
|
value445NormalizedList = list(normalized([float(x[0]) for x in valueList]))
|
|
|
value565NormalizedList = list(normalized([float(x[1]) for x in valueList]))
|
|
|
value670NormalizedList = list(normalized([float(x[2]) for x in valueList]))
|
|
|
# print(value445NormalizedList)
|
|
|
# 第二步，归一化的值转化为255的16进制
|
|
|
valueRList = [
|
|
|
hex(int(x * 255))[2:].zfill(2) if not np.isnan(x) else "ff"
|
|
|
for x in value445NormalizedList
|
|
|
]
|
|
|
# print(valueRList)
|
|
|
# print(hex(int(value445NormalizedList[0] * 255))[2:].zfill(2))
|
|
|
valueGList = [
|
|
|
hex(int(x * 255))[2:].zfill(2) if not np.isnan(x) else "ff"
|
|
|
for x in value565NormalizedList
|
|
|
]
|
|
|
valueBList = [
|
|
|
hex(int(x * 255))[2:].zfill(2) if not np.isnan(x) else "ff"
|
|
|
for x in value670NormalizedList
|
|
|
]
|
|
|
|
|
|
return ["#" + "".join(item) for item in zip(valueRList, valueGList, valueBList)]
|