因缺乏可重复的例子而提前道歉。我的数据结构的方式使这很困难。
我正在构建程序以帮助我收集数据(在本例中为网格坐标)以训练分类算法。我希望能够从我点击的情节中拉出特定的网格单元格。到目前为止,我使用fig.canvas.mpl_connect('button_press_event', onclick)
已经能够使用event.x
和event.y
提取像素坐标(我也尝试使用event.xdata
和{{1}但这似乎给了我一些我不认识的坐标?)但我无法提取我想要的实际数据。
在我的代码中,我使用以下模块:
event.ydata
我的数据来自存储在NetCDF文件中的数据,NetCDF文件是大型模型运行的子集,变量存储为165 * 116阵列(这是网格尺寸)。为了构建我的绘图,我使用以下函数:
from netCDF4 import Dataset # reads netCDF file
from os import listdir
from os.path import isfile, join
import matplotlib.pyplot as plt
import pylab
from mpl_toolkits.basemap import Basemap # basemap tools
from datetime import datetime, timedelta # for working with datetimes
from random import randint
import numpy as np
当使用正确的数据调用时会产生如下情节:
现在,我开始感到困惑。我从随机文件的沙子时间制作我的图,并尝试让它们返回我点击的数据并将其保存到数据库中。这是调用上述函数的代码:
def grab_sst_time(time_idx):
"""
This reads the datetime value in my NetCDF files into local time
"""
dtcon_days = time[time_idx]
dtcon_start = datetime(1990,1,1) # This is the "days since" part
dtcon_delta = timedelta(dtcon_days/24/60/60) # Create a time delta object from the number of days
dtcon_offset = dtcon_start + dtcon_delta # Add the specified number of days to 1990
frame_time = dtcon_offset
return frame_time
def plot_temp(temp, time_idx, fig_no):
"""
Plot temperature
"""
# make frame_idx an integer to avoid slicing errors
frame_idx = int(time_idx)
# get 'frame_time'
frame_time = grab_sst_time(frame_idx)
# map setup
fig = plt.figure()
fig.subplots_adjust(left=0., right=1., bottom=0., top=0.9)
# Setup the map
m = Basemap(projection='merc', llcrnrlat=-38.050653, urcrnrlat=-34.453367,\
llcrnrlon=147.996456, urcrnrlon=152.457344, lat_ts=20, resolution='h')
# draw stuff
m.drawcoastlines()
m.fillcontinents(color='black')
# plot temp
cs = m.pcolor(lons,lats,np.squeeze(temp), latlon = True ,vmin=temp_min, vmax=temp_max, cmap='plasma')
# plot colourbar
plt.colorbar()
# datetime title
plt.title('Regional - Temperature (Celcius)\n' +
frame_time.strftime("%Y-%m-%d %H:%M:%S") + ' | ' + str(fname) + '_idx: ' + str(frame_idx))
# stop axis from being cropped
plt.tight_layout()
return fig
这似乎工作正常,但我无法提取我真正想要的数据(NetCDF数据)。理想情况下,我想返回网格单元格(即165 * 116网格的坐标,类似于x = 112,y = 154)。而是点击给我一个如下输出:
def onclick(event):
"""
On click function for selecting data
"""
global ix, iy
ix, iy = event.xdata, event.ydata
print('x='+str(ix), 'y='+str(iy), 'class='+str(class_value), time_value)
global train_data
train_data.append((ix, iy, class, time_value))
# set colour scale variables
temp_min = 14
temp_max = 24
# list to hold data collected
train_data = []
# __Random Data__
# get list of files in data directory
directory = "/Users/directory/with/my/data"
file_ls = [f for f in listdir(directory) if isfile(join(directory, f))]
file_ls = list(filter(lambda x:'naroom_avg' in x, file_ls))
# set randomness seed
plot_num = 2
np.random.seed(1010)
rnd_file = np.random.randint(len(file_ls), size=plot_num)
rnd_times = np.random.randint(29, size=plot_num)
# __Make plots__
for i in range(0, plot_num):
# grab file
file_no = rnd_file[i]
file_path = directory + "/" + file_ls[file_no]
fname = str(file_ls[i])[11:16]
# grab time
time_idx = rnd_times[i]
fh = Dataset(file_path, mode='r')
# extract data
lats = fh.variables['lat_rho'][:]
lons = fh.variables['lon_rho'][:]
time = fh.variables['ocean_time'][:]
temp = fh.variables['temp'][time_idx,29,:,:]
# time output
time_value = grab_sst_time(time_idx)
# make interactive plot
fig = plot_temp(temp, time_idx, i)
cid = fig.canvas.mpl_connect('button_press_event', onclick)
# Select class A
class = 'A'
print('Select data...')
input("Press Enter to continue...")
plt.show()
有关如何获得所需输出的任何想法?
以下是我的NetCDF文件的相关变量的结构:
x=432047.683555 y=449210.017634 class=A 2004-06-17 12:00:00
x=448214.625063 y=430733.513053 class=A 2004-06-17 12:00:00
x=448214.625063 y=408792.663863 class=A 2004-06-17 12:00:00
x=448792.015832 y=441703.937648 class=A 2004-06-17 12:00:00
答案 0 :(得分:2)
要从lon,lat转换为投影坐标,请使用int value;
for (int a = 0; a < count; ++a)
{
InFile >> value;
cout << value << '\n';
}
值调用底图实例,
lon,lat
要将投影坐标转换回lon,lat坐标,请使用反向
m = Basemap(...)
xpt,ypt = m(lon,lat)
我们可能会修改here中的示例,以在屏幕上显示点击的坐标。
lon,lat = m(xpt,ypt, inverse=True)