Switch PyMC3 to PyMC (which is the new name starting from PyMC v4.0.0)

This MR updates the PyMC3 interface to use the new version of PyMC, which is now v4.0.0 (the version number suffix has been dropped). This is mainly a very straightforward port by replacing imports of pymc3 with pymc. The Theano library has also been renamed to aesara, so those are changed too.

The test suite runs and also the two examples, linear_regression_pymc.py and linear_regression_pymc_custom_likelihood.py, both work exactly as expected.

test/integration/sampler_run_test.py

+import multiprocessing
+multiprocessing.set_start_method("fork")  # noqa
+
 import unittest
 import pytest
 import shutil
@@ -33,20 +36,23 @@ class TestRunningSamplers(unittest.TestCase):
         bilby.core.utils.command_line_args.bilby_test_mode = False
         shutil.rmtree("outdir")
 
+    def _run_sampler(self, sampler, **kwargs):
+        kwargs["resume"] = kwargs.get("resume", False)
+        for npool in [1, 2]:
+            _ = bilby.run_sampler(
+                likelihood=self.likelihood,
+                priors=self.priors,
+                sampler=sampler,
+                save=False,
+                npool=npool,
+                **kwargs,
+            )
+
     def test_run_cpnest(self):
-        _ = bilby.run_sampler(
-            likelihood=self.likelihood,
-            priors=self.priors,
-            sampler="cpnest",
-            nlive=100,
-            save=False,
-            resume=False,
-        )
+        self._run_sampler(sampler="cpnest", nlive=100)
 
     def test_run_dnest4(self):
-        _ = bilby.run_sampler(
-            likelihood=self.likelihood,
-            priors=self.priors,
+        self._run_sampler(
             sampler="dnest4",
             max_num_levels=2,
             num_steps=10,
@@ -54,22 +60,13 @@ class TestRunningSamplers(unittest.TestCase):
             num_per_step=10,
             thread_steps=1,
             num_particles=50,
-            save=False,
         )
 
     def test_run_dynesty(self):
-        _ = bilby.run_sampler(
-            likelihood=self.likelihood,
-            priors=self.priors,
-            sampler="dynesty",
-            nlive=100,
-            save=False,
-        )
+        self._run_sampler(sampler="dynesty", nlive=100)
 
     def test_run_dynamic_dynesty(self):
-        _ = bilby.run_sampler(
-            likelihood=self.likelihood,
-            priors=self.priors,
+        self._run_sampler(
             sampler="dynamic_dynesty",
             nlive_init=100,
             nlive_batch=100,
@@ -77,124 +74,73 @@ class TestRunningSamplers(unittest.TestCase):
             maxbatch=0,
             maxcall=100,
             bound="single",
-            save=False,
         )
 
     def test_run_emcee(self):
-        _ = bilby.run_sampler(
-            likelihood=self.likelihood,
-            priors=self.priors,
-            sampler="emcee",
-            iterations=1000,
-            nwalkers=10,
-            save=False,
-        )
+        self._run_sampler(sampler="emcee", iterations=1000, nwalkers=10)
 
     def test_run_kombine(self):
-        _ = bilby.run_sampler(
-            likelihood=self.likelihood,
-            priors=self.priors,
+        self._run_sampler(
             sampler="kombine",
             iterations=2000,
             nwalkers=20,
-            save=False,
             autoburnin=False,
         )
 
     def test_run_nestle(self):
-        _ = bilby.run_sampler(
-            likelihood=self.likelihood,
-            priors=self.priors,
-            sampler="nestle",
-            nlive=100,
-            save=False,
-        )
+        self._run_sampler(sampler="nestle", nlive=100)
 
     def test_run_nessai(self):
-        _ = bilby.run_sampler(
-            likelihood=self.likelihood,
-            priors=self.priors,
+        self._run_sampler(
             sampler="nessai",
             nlive=100,
             poolsize=1000,
             max_iteration=1000,
-            save=False,
         )
 
     def test_run_pypolychord(self):
         pytest.importorskip("pypolychord")
-        _ = bilby.run_sampler(
-            likelihood=self.likelihood,
-            priors=self.priors,
-            sampler="pypolychord",
-            nlive=100,
-            save=False,
-        )
+        self._run_sampler(sampler="pypolychord", nlive=100)
 
     def test_run_ptemcee(self):
-        _ = bilby.run_sampler(
-            likelihood=self.likelihood,
-            priors=self.priors,
+        self._run_sampler(
             sampler="ptemcee",
             nsamples=100,
             nwalkers=50,
             burn_in_act=1,
             ntemps=1,
             frac_threshold=0.5,
-            save=False,
         )
 
     @pytest.mark.skipif(sys.version_info[1] <= 6, reason="pymc3 is broken in py36")
     def test_run_pymc3(self):
-        _ = bilby.run_sampler(
-            likelihood=self.likelihood,
-            priors=self.priors,
+        self._run_sampler(
             sampler="pymc3",
             draws=50,
             tune=50,
             n_init=250,
-            save=False,
         )
 
     def test_run_pymultinest(self):
-        _ = bilby.run_sampler(
-            likelihood=self.likelihood,
-            priors=self.priors,
-            sampler="pymultinest",
-            nlive=100,
-            save=False,
-        )
+        self._run_sampler(sampler="pymultinest", nlive=100)
 
     def test_run_PTMCMCSampler(self):
-        _ = bilby.run_sampler(
-            likelihood=self.likelihood,
-            priors=self.priors,
+        self._run_sampler(
             sampler="PTMCMCsampler",
             Niter=101,
             burn=2,
             isave=100,
-            save=False,
         )
 
     def test_run_ultranest(self):
         # run using NestedSampler (with nlive specified)
-        _ = bilby.run_sampler(
-            likelihood=self.likelihood, priors=self.priors,
-            sampler="ultranest", nlive=100, save=False,
-        )
+        self._run_sampler(sampler="ultranest", nlive=100, resume="overwrite")
 
         # run using ReactiveNestedSampler (with no nlive given)
-        _ = bilby.run_sampler(
-            likelihood=self.likelihood, priors=self.priors,
-            sampler='ultranest', save=False,
-        )
+        self._run_sampler(sampler='ultranest', resume="overwrite")
 
     def test_run_bilby_mcmc(self):
-        _ = bilby.run_sampler(
-            likelihood=self.likelihood, priors=self.priors,
-            sampler="bilby_mcmc", nsamples=200, save=False,
-            printdt=1,
-        )
+        self._run_sampler(sampler="bilby_mcmc", nsamples=200, printdt=1)
 
 
 if __name__ == "__main__":