dirt
commented
6 months ago Jiapei, thanks for the feedback and information. To better pinpoint the problem, would you please help to share a sample ndvi array for you to debug? You can send it to @.***
Kaiguang
Get Outlook for iOShttps://aka.ms/o0ukef
From: Jiapei Ma @.> Sent: Tuesday, December 12, 2023 9:09:02 PM To: zhaokg/Rbeast @.> Cc: Subscribed @.***> Subject: [zhaokg/Rbeast] calculations are incomplete in beast123 when using large array as input (Issue #22)
I found that beast123 can't calculate compeletly when the input array is big enough(number of row or column is bigger than 200) `import numpy as np import Rbeast as rb import xarray as xr import matplotlib. pyplot as plt ndvi3d = np. load("sub_r_ndvi. npz")['r_nvdi']
I found that beast123 can't calculate compeletly when the input array is big enough(number of row or column is bigger than 200)
`import numpy as np import Rbeast as rb import xarray as xr import matplotlib.pyplot as plt
ndvi3d = np.load("sub_r_ndvi.npz")['r_nvdi'] # (722, 1378, 38) array
metadata = lambda: None metadata.start = 1986 metadata.season = 'none' metadata.whichDimIsTime = 3 metadata.period = 1.0 extra = rb.args() extra.numThreadsPerCPU = 2 o = rb.beast123(ndvi3d, metadata, [], [], extra) plt.imshow(o.trend.cp[:, :, 0]) plt.show()`
the abolve code shows a plot like
however when we decrease the number of row and column ndvi3d = ndvi3d[0:150, 0:150, :] the calculation can be completely like 8140.jpg (view on web)https://urldefense.com/v3/__https://github.com/zhaokg/Rbeast/assets/26560891/627a7b2c-9e3b-4c9d-91f8-d6b134c79d5e__;!!KGKeukY!0f9Mzvv8lRMkzBlkuztO3yMuhFa8LE2dmzQFC4t2_52_-t2GVQP7yy0PiqBj2YEQv7as2Yv2UaimBIhj33V0dis0DRBQ$
— Reply to this email directly, view it on GitHubhttps://urldefense.com/v3/__https://github.com/zhaokg/Rbeast/issues/22__;!!KGKeukY!0f9Mzvv8lRMkzBlkuztO3yMuhFa8LE2dmzQFC4t2_52_-t2GVQP7yy0PiqBj2YEQv7as2Yv2UaimBIhj33V0dqFLMRAs$, or unsubscribehttps://urldefense.com/v3/__https://github.com/notifications/unsubscribe-auth/AMCAGWQB4HEXXFMNMJHQICLYJEZ65AVCNFSM6AAAAABASRTG72VHI2DSMVQWIX3LMV43ASLTON2WKOZSGAZTQOJWGUZTANQ__;!!KGKeukY!0f9Mzvv8lRMkzBlkuztO3yMuhFa8LE2dmzQFC4t2_52_-t2GVQP7yy0PiqBj2YEQv7as2Yv2UaimBIhj33V0dqPr9rw-$. You are receiving this because you are subscribed to this thread.Message ID: @.***>
MaJiapei
zhaokg
I found that beast123 can't calculate compeletly when the input array is big enough(number of row or column is bigger than 200)
`import numpy as np import Rbeast as rb import xarray as xr import matplotlib.pyplot as plt
ndvi3d = np.load("sub_r_ndvi.npz")['r_nvdi'] # (722, 1378, 38) array
metadata = lambda: None metadata.start = 1986 metadata.season = 'none' metadata.whichDimIsTime = 3 metadata.period = 1.0 extra = rb.args() extra.numThreadsPerCPU = 2 o = rb.beast123(ndvi3d, metadata, [], [], extra) plt.imshow(o.trend.cp[:, :, 0]) plt.show()`
the abolve code shows a plot like
however when we decrease the number of row and column
ndvi3d = ndvi3d[0:150, 0:150, :]the calculation can be completely like