Performance test for LinearRegression in mn4 fails

[salvisolamartinell]

Describe the bug The performance test for LinearRegression in mn4 is failing. That is using the kdd99 dataset (/gpfs/projects/bsc19/COMPSs_DATASETS/dislib/kdd99/train.csv). The task _compute_model_parameters fails with a Singular Matrix error. I don't know if this test ever worked. This is the stacktrace:

2020-10-08 13:29:45,783 ERROR pycompss.worker.piper.piper_worker worker - Traceback (most recent call last):
  File "/apps/COMPSs/2.7/Bindings/python/3/pycompss/worker/commons/worker.py", line 268, in task_execution
    **compss_kwargs)
  File "/apps/COMPSs/2.7/Bindings/python/3/pycompss/api/task.py", line 267, in task_decorator
    return self.worker_call(*args, **kwargs)
  File "/apps/COMPSs/2.7/Bindings/python/3/pycompss/api/task.py", line 1492, in worker_call
    kwargs['compss_tracing'])
  File "/apps/COMPSs/2.7/Bindings/python/3/pycompss/api/task.py", line 1843, in execute_user_code
    user_returns = self.user_function(*user_args, **user_kwargs)
  File "/home/bsc19/bsc19029/git/dislib_no_outs/dislib/regression/linear/base.py", line 191, in _compute_model_parameters
    params = np.linalg.solve(ztz, zty)
  File "<__array_function__ internals>", line 6, in solve
  File "/apps/PYTHON/3.6.4_ML/lib/python3.6/site-packages/numpy/linalg/linalg.py", line 399, in solve
    r = gufunc(a, b, signature=signature, extobj=extobj)
  File "/apps/PYTHON/3.6.4_ML/lib/python3.6/site-packages/numpy/linalg/linalg.py", line 97, in _raise_linalgerror_singular
    raise LinAlgError("Singular matrix")
numpy.linalg.LinAlgError: Singular matrix

To Reproduce In MN4 run the performance test /gpfs/projects/bsc19/PERFORMANCE/dislib/tests/lr.py with COMPSs. I had to install the cvxpy dependency locally before.

Environment:

• SC: Mn4
• COMPSs version: 2.7
• Dislib version: master
• Java / Python versions: java 8 / python 3.6.4_ML