TypeError：图像数据无法转换为float - 使用NcML数组

Question

我正在尝试使用NcML数组运行NDVI更改分析，我使用xarray和opendap取消http://www.auscover.org.au/purl/lpdaac-mosaic-mod13q1-v5。

我已经切割了我需要的数据并将它们分配给火灾变量之前，期间和之后。

现在，当我尝试在同一图中显示三个图时，我收到一个错误：'图像数据无法转换为浮动'。我错过了什么吗？我认为我分配的数组是xml而不是图像？

任何建议都会受到赞赏，因为这份报告将在明天发布。

import xarray as xr
import numpy as np
import pandas as pd

import matplotlib.pyplot as plt
import seaborn
%matplotlib inline
seaborn.set_style('dark')

NDVI_aggr_data_url = 'http://data.auscover.org.au/thredds/dodsC/auscover/lpdaac-aggregates/c5/v2-nc4/aust/MOD13Q1.005/MOD13Q1.aggregated.aust.005.normalised_difference_vegetation_index.ncml'

NDVI_aggr = xr.open_dataset(NDVI_aggr_data_url)
NDVI_aggr

lat_bounds = slice(-36.341, -36.645)
lon_bounds = slice(146.666, 147.133)

time_bounds = slice('2017-02-08', '2017-02-20')

beechworth_NDVI_post = NDVI_aggr.sel(
    latitude=lat_bounds, longitude=lon_bounds, time=time_bounds)
beechworth_NDVI_post

beechworth_NDVI_post.load()
plt.rcParams["figure.figsize"] = (12,10)

beechworth_NDVI_post.ndvi.plot.imshow(col='time', cmap='viridis')
plt.title('NDVI - 18 February 2017', y=1.1)
plt.xlabel('Longitude')
plt.ylabel('Latitude')
plt.savefig("2017_NDVI_test1.png", dpi=100)

post = beechworth_NDVI_post

lat_bounds = slice(-36.341, -36.645)
lon_bounds = slice(146.666, 147.133)

time_bounds = slice('2009-02-07', '2009-02-20')

beechworth_NDVI_during = NDVI_aggr.sel(
    latitude=lat_bounds, longitude=lon_bounds, time=time_bounds)
beechworth_NDVI_during
beechworth_NDVI_during.load()
plt.rcParams["figure.figsize"] = (12,10)

beechworth_NDVI_during.ndvi.plot.imshow(col='time', cmap='viridis')
plt.title('NDVI - 18 February 2009', y=1.1)
plt.xlabel('Longitude')
plt.ylabel('Latitude')
plt.savefig("2009_NDVI.png", dpi=100)

during = beechworth_NDVI_during

lat_bounds = slice(-36.341, -36.645)
lon_bounds = slice(146.666, 147.133)

time_bounds = slice('2008-02-07', '2008-02-20')

beechworth_NDVI_before = NDVI_aggr.sel(
    latitude=lat_bounds, longitude=lon_bounds, time=time_bounds)
beechworth_NDVI_before

beechworth_NDVI_before.load()
plt.rcParams["figure.figsize"] = (12,10)

beechworth_NDVI_before.ndvi.plot.imshow(col='time', cmap='viridis')
plt.title('NDVI - 18 February 2008', y=1.1)
plt.xlabel('Longitude')
plt.ylabel('Latitude')
plt.savefig("2008_NDVI.png", dpi=100)

before = beechworth_NDVI_before

figure, ax_s = plt.subplots(ncols=3)
plt.title('NDVI in Beechworth before, during, and after a bushfire')
for data, ax in zip([before, during, post], ax_s):
    ax.imshow(data, cmap='viridis', vmin=0, vmax=0.9)

Answer 1

看起来您的数据子设置有几个问题。这是一个例子，使用适当的选择。

before = beechworth_NDVI_before.isel(time=0,nv=1).ndvi
during = beechworth_NDVI_during.isel(time=0,nv=1).ndvi
post = beechworth_NDVI_post.isel(time=0,nv=1).ndvi


figure, ax_s = plt.subplots(ncols=3)
plt.title('NDVI in Beechworth before, during, and after a bushfire')
for data, ax in zip([before, during, post], ax_s):
    ax.imshow(data, cmap='viridis', vmin=0, vmax=0.9)

导致