PwBaseWorkChain: add handler for intentional non-convergence (#695)

In certain cases, a user will set inputs such that the SCF will not
converge, for example, by setting `scf_must_converge` to false, or by
setting `electron_maxstep` to zero. This can be useful to initialize a
charge density of a system and the workchain should not restart or fail
in this case.

The `PwParser` now returns `WARNING_ELECTRONIC_CONVERGENCE_NOT_REACHED`
instead of `ERROR_ELECTRONIC_CONVERGENCE_NOT_REACHED` if any of these
parameters are set as described. The `PwBaseWorkChain` has a new handler
for this new exit code to consider the calculation as successful and
finish the workchain.

Co-authored-by: Sebastiaan Huber <mail@sphuber.net>

aiida_quantumespresso/calculations/pw.py

@@ -145,6 +145,11 @@ def define(cls, spec):
  message='The electronic minimization cycle did not reach self-consistency.')
  spec.exit_code(541, 'ERROR_SYMMETRY_NON_ORTHOGONAL_OPERATION',
  message='The variable cell optimization broke the symmetry of the k-points.')
+
+ # Strong warnings about calculation results, but something tells us that you're ok with that
+ spec.exit_code(710, 'WARNING_ELECTRONIC_CONVERGENCE_NOT_REACHED',
+ message='The electronic minimization cycle did not reach self-consistency, but `scf_must_converge` '
+ 'is `False` and/or `electron_maxstep` is 0.')
  # yapf: enable
 
  @classproperty

aiida_quantumespresso/parsers/pw.py

@@ -150,7 +150,14 @@ def validate_electronic(self, trajectory, parameters, logs):
  return
 
  if 'ERROR_ELECTRONIC_CONVERGENCE_NOT_REACHED' in logs['error']:
- return self.exit_codes.ERROR_ELECTRONIC_CONVERGENCE_NOT_REACHED
+ scf_must_converge = self.node.inputs.parameters.get_attribute('ELECTRONS',
+ {}).get('scf_must_converge', True)
+ electron_maxstep = self.node.inputs.parameters.get_attribute('ELECTRONS', {}).get('electron_maxstep', 1)
+
+ if electron_maxstep == 0 or not scf_must_converge:
+ return self.exit_codes.WARNING_ELECTRONIC_CONVERGENCE_NOT_REACHED
+ else:
+ return self.exit_codes.ERROR_ELECTRONIC_CONVERGENCE_NOT_REACHED
 
  def validate_dynamics(self, trajectory, parameters, logs):
  """Analyze problems that are specific to `dynamics` type calculations: i.e. `md` and `vc-md`."""

aiida_quantumespresso/workflows/pw/base.py

@@ -115,6 +115,9 @@ def define(cls, spec):
  message='The initialization calculation failed.')
  spec.exit_code(501, 'ERROR_IONIC_CONVERGENCE_REACHED_EXCEPT_IN_FINAL_SCF',
  message='Then ionic minimization cycle converged but the thresholds are exceeded in the final SCF.')
+ spec.exit_code(710, 'WARNING_ELECTRONIC_CONVERGENCE_NOT_REACHED',
+ message='The electronic minimization cycle did not reach self-consistency, but `scf_must_converge` '
+ 'is `False` and/or `electron_maxstep` is 0.')
  # yapf: enable
 
  @classmethod
@@ -588,3 +591,15 @@ def handle_electronic_convergence_not_achieved(self, calculation):
  action = f'reduced beta mixing from {mixing_beta} to {mixing_beta_new} and restarting from the last calculation'
  self.report_error_handled(calculation, action)
  return ProcessHandlerReport(True)
+
+ @process_handler(priority=420, exit_codes=[
+ PwCalculation.exit_codes.WARNING_ELECTRONIC_CONVERGENCE_NOT_REACHED,
+ ])
+ def handle_electronic_convergence_warning(self, calculation):
+ """Handle `WARNING_ELECTRONIC_CONVERGENCE_NOT_REACHED': consider finished."""
+ self.ctx.is_finished = True
+ self.ctx.restart_calc = calculation
+ action = 'electronic convergence not reached but inputs say this is ok: consider finished.'
+ self.report_error_handled(calculation, action)
+ self.results() # Call the results method to attach the output nodes
+ return ProcessHandlerReport(True, self.exit_codes.WARNING_ELECTRONIC_CONVERGENCE_NOT_REACHED)

tests/parsers/fixtures/pw/failed_scf_not_converged_intentional/aiida.out

@@ -0,0 +1,167 @@
+
+ Program PWSCF v.6.1 (svn rev. 13369) starts on 7Jun2019 at 17:14: 0
+
+ This program is part of the open-source Quantum ESPRESSO suite
+ for quantum simulation of materials; please cite
+ "P. Giannozzi et al., J. Phys.:Condens. Matter 21 395502 (2009);
+ URL http://www.quantum-espresso.org",
+ in publications or presentations arising from this work. More details at
+ http://www.quantum-espresso.org/quote
+
+ Parallel version (MPI), running on 1 processors
+ Reading input from aiida.in
+
+ Current dimensions of program PWSCF are:
+ Max number of different atomic species (ntypx) = 10
+ Max number of k-points (npk) = 40000
+ Max angular momentum in pseudopotentials (lmaxx) = 3
+
+ Subspace diagonalization in iterative solution of the eigenvalue problem:
+ a serial algorithm will be used
+
+
+ G-vector sticks info
+ --------------------
+ sticks: dense smooth PW G-vecs: dense smooth PW
+ Sum 859 433 127 16889 5985 965
+
+
+
+ bravais-lattice index = 0
+ lattice parameter (alat) = 7.2558 a.u.
+ unit-cell volume = 270.1072 (a.u.)^3
+ number of atoms/cell = 2
+ number of atomic types = 1
+ number of electrons = 8.00
+ number of Kohn-Sham states= 4
+ kinetic-energy cutoff = 30.0000 Ry
+ charge density cutoff = 240.0000 Ry
+ convergence threshold = 1.0E-06
+ mixing beta = 0.7000
+ number of iterations used = 8 plain mixing
+ Exchange-correlation = PBE ( 1 4 3 4 0 0)
+
+ celldm(1)= 7.255773 celldm(2)= 0.000000 celldm(3)= 0.000000
+ celldm(4)= 0.000000 celldm(5)= 0.000000 celldm(6)= 0.000000
+
+ crystal axes: (cart. coord. in units of alat)
+ a(1) = ( 0.707107 0.707107 0.000000 )
+ a(2) = ( 0.707107 0.000000 0.707107 )
+ a(3) = ( 0.000000 0.707107 0.707107 )
+
+ reciprocal axes: (cart. coord. in units 2 pi/alat)
+ b(1) = ( 0.707107 0.707107 -0.707107 )
+ b(2) = ( 0.707107 -0.707107 0.707107 )
+ b(3) = ( -0.707107 0.707107 0.707107 )
+
+
+ PseudoPot. # 1 for Si read from file:
+ ./pseudo/Si.pbe-n-rrkjus_psl.1.0.0.UPF
+ MD5 check sum: 0b0bb1205258b0d07b9f9672cf965d36
+ Pseudo is Ultrasoft + core correction, Zval = 4.0
+ Generated using "atomic" code by A. Dal Corso v.5.1
+ Using radial grid of 1141 points, 6 beta functions with:
+ l(1) = 0
+ l(2) = 0
+ l(3) = 1
+ l(4) = 1
+ l(5) = 2
+ l(6) = 2
+ Q(r) pseudized with 0 coefficients
+
+
+ atomic species valence mass pseudopotential
+ Si 4.00 28.08550 Si( 1.00)
+
+
+---- OUTPUT REMOVED ----
+
+ Cartesian axes
+
+ site n. atom positions (alat units)
+ 1 Si tau( 1) = ( 0.0000000 0.0000000 0.0000000 )
+ 2 Si tau( 2) = ( 0.3535534 0.3535534 0.3535534 )
+
+ Crystallographic axes
+
+ site n. atom positions (cryst. coord.)
+ 1 Si tau( 1) = ( 0.0000000 0.0000000 0.0000000 )
+ 2 Si tau( 2) = ( 0.2500000 0.2500000 0.2500000 )
+
+ number of k points= 3
+ cart. coord. in units 2pi/alat
+ k( 1) = ( 0.0000000 0.0000000 0.0000000), wk = 0.2500000
+ k( 2) = ( 0.3535534 -0.3535534 -0.3535534), wk = 1.0000000
+ k( 3) = ( 0.0000000 0.0000000 -0.7071068), wk = 0.7500000
+
+ cryst. coord.
+ k( 1) = ( 0.0000000 0.0000000 0.0000000), wk = 0.2500000
+ k( 2) = ( 0.0000000 0.0000000 -0.5000000), wk = 1.0000000
+ k( 3) = ( 0.0000000 -0.5000000 -0.5000000), wk = 0.7500000
+
+ Dense grid: 16889 G-vectors FFT dimensions: ( 36, 36, 36)
+
+ Smooth grid: 5985 G-vectors FFT dimensions: ( 25, 25, 25)
+
+ Estimated max dynamical RAM per process > 10.86MB
+
+ Initial potential from superposition of free atoms
+
+ starting charge 7.99888, renormalised to 8.00000
+ Starting wfc are 8 randomized atomic wfcs
+
+ total cpu time spent up to now is 0.9 secs
+
+ per-process dynamical memory: 20.0 Mb
+
+ Self-consistent Calculation
+
+ iteration # 1 ecut= 30.00 Ry beta=0.70
+ Davidson diagonalization with overlap
+ ethr = 1.00E-02, avg # of iterations = 2.0
+
+ total cpu time spent up to now is 1.1 secs
+
+ total energy = -22.64340821 Ry
+ Harris-Foulkes estimate = -22.67223092 Ry
+ estimated scf accuracy < 0.10529730 Ry
+
+ iteration # 2 ecut= 30.00 Ry beta=0.70
+ Davidson diagonalization with overlap
+ ethr = 1.32E-03, avg # of iterations = 1.0
+
+ total cpu time spent up to now is 1.3 secs
+
+ total energy = -22.64972429 Ry
+ Harris-Foulkes estimate = -22.65005091 Ry
+ estimated scf accuracy < 0.00535578 Ry
+
+ iteration # 3 ecut= 30.00 Ry beta=0.70
+ Davidson diagonalization with overlap
+ ethr = 6.69E-05, avg # of iterations = 3.0
+
+ total cpu time spent up to now is 1.5 secs
+
+ total energy = -22.65168183 Ry
+ Harris-Foulkes estimate = -22.65176063 Ry
+ estimated scf accuracy < 0.00032274 Ry
+
+ End of self-consistent calculation
+
+ convergence NOT achieved after 3 iterations: stopping
+
+ Writing output data file aiida.save
+
+---- OUTPUT REMOVED ----
+
+ Parallel routines
+ fft_scatter : 0.01s CPU 0.01s WALL ( 343 calls)
+
+ PWSCF : 1.37s CPU 1.51s WALL
+
+
+ This run was terminated on: 17:14: 2 7Jun2019
+
+=------------------------------------------------------------------------------=
+ JOB DONE.
+=------------------------------------------------------------------------------=