search

ManonDls / KFTS-InSAR

Kalman Filter Time series Analysis for InSAR
GNU General Public License v3.0
54 stars 11 forks source link

H5py version error #1

Closed Arctic-Ambrun closed 3 years ago

Arctic-Ambrun commented 3 years ago

Hi, I tried to use the test data, however, an bug occurred and Here is my result. I suspect that caused by the different version of H5PY.

In my PC, the version of required wheels are:

h5py 2.10.0 py36hd6299e0_1
hdf5 1.10.6 hb1b8bf9_0
matplotlib 3.3.4 py36h06a4308_0
matplotlib-base 3.3.4 py36h62a2d02_0
future 0.18.2 py36_1
scipy 1.5.2 py36h0b6359f_0

Thank you!

Screenshot from 2021-09-09 13-06-50

ManonDls commented 3 years ago

Hello, thank you for rising this issue. The problem was the systematic use of MPI... As the test does not require multiprocessing, I disabled the use of MPI in such configuration. To use multiprocessing, h5py has to be installed with the openmpi and gcc8 dependency so that there is no conflict in writing in the same file at the same time (I updated the requirement specifications in README.md)

Arctic-Ambrun commented 3 years ago

Thank you for your rapid response, really appreciate Now seems the H5PY problem has been solved. Still there are some error in the name of the test data. I suppose the name of test file has been renamed from States.h5 to Etna_sample.h5? So the exist OS error.

Screenshot from 2021-09-10 17-18-45

ManonDls commented 3 years ago

Sorry the empty KFTS folder in testdata is missing, you can create it

Arctic-Ambrun commented 3 years ago

Thank you very much! Still there are file missing, when processing the last command. python plotoutput.py -c configs/config_Etna.ini -geom ./ -rmsTh 2 It shows that there lack "RMS_map.h5" file in the /testdata/KFTS/RMS_map.h5

ManonDls commented 3 years ago

If the command python kf2rms.py -c configs/config_Etna.ini runs properly, it creates this file containing Root Mean Square errors.