Issues about parallel computation

[Devilbean000]

Describe the bug A clear and concise description of what the bug is.

Two issues:

1. When I turn on parallel parameter, I get TypeError.
2. I get different results when I turn on and off the parallel computation. Does parallel computation influence results?

To Reproduce Short code showing how to reproduce the issue. S = f t = np.arange(1970,2020,1)

EEMD options

max_imf=3

Prepare and run EEMD

eemd = EEMD(trials=10000, FIXE_H=10, DTYPE=np.float16, noise_width=0.2, parallel=True) #max_imfs=10) eemd.noise_seed(6021)

Execute EEMD on S

eIMFs = eemd.eemd(S, t, max_imf)

Expected behavior A clear and concise description of what you expected to happen. I get TypeError: 'Pool' object is not callable

Details: 'TypeError Traceback (most recent call last) Cell In[4], line 12 9 eemd.noise_seed(6021) 11 # Execute EEMD on S ---> 12 eIMFs = eemd.eemd(S, t, max_imf)

File ~/.conda/envs/eemd/lib/python3.11/site-packages/PyEMD/EEMD.py:189, in EEMD.eemd(self, S, T, max_imf, progress) 187 else: 188 map_pool = map --> 189 all_IMFs = map_pool(self._trial_update, range(self.trials)) 191 if self.parallel: 192 map_pool.close()'

Screenshots If applicable, add screenshots to help explain your problem.

Running environment Provide operating system (OS) information, PyEMD version and describe virtual environment (if any). Python 3.11.4, latest PyEMD

Additional context Add any other context about the problem here. I used parallel computation before, now I have new python and conda environment. Then I can't turn on the parallel parameter, therefore I turn it off. But it's strange that I get different results with same manuscript. Does parallel computation influence results?

[sebhinck]

Can be solved by changing https://github.com/laszukdawid/PyEMD/blob/7361a883e21bf92a9a62abfeaceffc38e91955b9/PyEMD/EEMD.py#L185-L192 to

        if self.parallel:
            pool  = Pool(processes=self.processes)
            map_pool = pool.map
        else:
            map_pool = map
        all_IMFs = map_pool(self._trial_update, range(self.trials))

        if self.parallel:
            pool.close()

[laszukdawid]

This should be fixed in the latest 1.5.2 version