kmeans takes too long

[euhruska]

timing data for analyse unit (run-tica-msm.py)

('time tica finished', '1151.17602015')
('time kmeans finished', '8621.65739107')
('time msm finished', '9160.20796895')
('time frame selection finished', '9189.05303311')
('time writing new frames finished', '9208.67641306')
('time plotting finished', '9623.44697213')

launched with

export PYEMMA_NJOBS=1
export OMP_NUM_THREADS=1 
/opt/xalt/0.7.6/sles11.3/bin/aprun -n 1 -N 1 -L 18544 -d 1 -cc 0 python "run-tica-msm.py" "--path" "/u/sciteam/hruska/scratch/extasy-tica" "--n_select" "100" "--cur_iter" "8" "--Kconfig" "settings_ala12_gpu_tica.wcfg" > "analyse.log"

The line cl = pyemma.coordinates.cluster_kmeans(data=y, k=msm_states, max_iter=10, stride=msm_stride) takes forever when having multiple iterations to analyse.

('n atoms', 132)
('n frames total', 4800000)
('n trajs', 900)

path /u/sciteam/hruska/scratch/radical.pilot.sandbox/rp.session.leonardo.rice.edu.eh22.017756.0001/

[euhruska]

tried

cl = pyemma.coordinates.cluster_mini_batch_kmeans(data=y, k=msm_states, max_iter=10, n_jobs=None)

didn't help

[euhruska]

trying http://scikit-learn.org/stable/modules/generated/sklearn.cluster.KMeans.html or https://github.com/radical-cybertools/midas/blob/master/k-means