Exemplo n.º 1
0
    def setup(self):
        """Set everything up..."""

        # check arguments are all ascii

        Debug.write("Start parsing command line: " + str(sys.argv))

        for token in sys.argv:
            try:
                token.encode("ascii")
            except UnicodeDecodeError:
                raise RuntimeError("non-ascii characters in input")

        self._argv = copy.deepcopy(sys.argv)

        replacements = {
            "-2d": "pipeline=2d",
            "-2di": "pipeline=2di",
            "-3d": "pipeline=3d",
            "-3di": "pipeline=3di",
            "-3dii": "pipeline=3dii",
            "-3dd": "pipeline=3dd",
            "-dials": "pipeline=dials",
            "-quick": "dials.fast_mode=true",
            "-failover": "failover=true",
            "-small_molecule": "small_molecule=true",
        }
        for k, v in replacements.iteritems():
            if k in self._argv:
                print(
                    "***\nCommand line option %s is deprecated.\nPlease use %s instead\n***"
                    % (k, v))
                self._argv[self._argv.index(k)] = v
        if "-atom" in self._argv:
            idx = self._argv.index("-atom")
            element = self._argv[idx + 1]
            self._argv[idx:idx + 2] = ["atom=%s" % element]
            print(
                "***\nCommand line option -atom %s is deprecated.\nPlease use atom=%s instead\n***"
                % (element, element))

        # first of all try to interpret arguments as phil parameters/files

        from xia2.Handlers.Phil import master_phil
        from libtbx.phil import command_line

        cmd_line = command_line.argument_interpreter(master_phil=master_phil)
        working_phil, self._argv = cmd_line.process_and_fetch(
            args=self._argv, custom_processor="collect_remaining")

        PhilIndex.merge_phil(working_phil)
        try:
            params = PhilIndex.get_python_object()
        except RuntimeError as e:
            raise Sorry(e)

        # sanity check / interpret Auto in input
        from libtbx import Auto

        if params.xia2.settings.input.atom is None:
            if params.xia2.settings.input.anomalous is Auto:
                PhilIndex.update("xia2.settings.input.anomalous=false")
        else:
            if params.xia2.settings.input.anomalous is False:
                raise Sorry(
                    "Setting anomalous=false and atom type inconsistent")
            params.xia2.settings.input.anomalous = True
            PhilIndex.update("xia2.settings.input.anomalous=true")

        if params.xia2.settings.resolution.keep_all_reflections is Auto:
            if (params.xia2.settings.small_molecule is True
                    and params.xia2.settings.resolution.d_min is None
                    and params.xia2.settings.resolution.d_max is None):
                PhilIndex.update(
                    "xia2.settings.resolution.keep_all_reflections=true")
            else:
                PhilIndex.update(
                    "xia2.settings.resolution.keep_all_reflections=false")

        if params.xia2.settings.small_molecule is True:
            Debug.write("Small molecule selected")
            if params.xia2.settings.symmetry.chirality is None:
                PhilIndex.update("xia2.settings.symmetry.chirality=nonchiral")
            params = PhilIndex.get_python_object()

        # pipeline options
        self._read_pipeline()

        for (parameter, value) in (
            ("project", params.xia2.settings.project),
            ("crystal", params.xia2.settings.crystal),
        ):
            validate_project_crystal_name(parameter, value)

        Debug.write("Project: %s" % params.xia2.settings.project)
        Debug.write("Crystal: %s" % params.xia2.settings.crystal)

        # FIXME add some consistency checks in here e.g. that there are
        # images assigned, there is a lattice assigned if cell constants
        # are given and so on

        params = PhilIndex.get_python_object()
        mp_params = params.xia2.settings.multiprocessing
        from xia2.Handlers.Environment import get_number_cpus

        if mp_params.mode == "parallel":
            if mp_params.type == "qsub":
                if which("qsub") is None:
                    raise Sorry("qsub not available")
            if mp_params.njob is Auto:
                mp_params.njob = get_number_cpus()
                if mp_params.nproc is Auto:
                    mp_params.nproc = 1
            elif mp_params.nproc is Auto:
                mp_params.nproc = get_number_cpus()
        elif mp_params.mode == "serial":
            if mp_params.type == "qsub":
                if which("qsub") is None:
                    raise Sorry("qsub not available")
            if mp_params.njob is Auto:
                mp_params.njob = 1
            if mp_params.nproc is Auto:
                mp_params.nproc = get_number_cpus()

        PhilIndex.update("xia2.settings.multiprocessing.njob=%d" %
                         mp_params.njob)
        PhilIndex.update("xia2.settings.multiprocessing.nproc=%d" %
                         mp_params.nproc)
        params = PhilIndex.get_python_object()
        mp_params = params.xia2.settings.multiprocessing

        if mp_params.nproc > 1 and os.name == "nt":
            raise Sorry("nproc > 1 is not supported on Windows.")  # #191

        if params.xia2.settings.indexer is not None:
            add_preference("indexer", params.xia2.settings.indexer)
        if params.xia2.settings.refiner is not None:
            add_preference("refiner", params.xia2.settings.refiner)
        if params.xia2.settings.integrater is not None:
            add_preference("integrater", params.xia2.settings.integrater)
        if params.xia2.settings.scaler is not None:
            add_preference("scaler", params.xia2.settings.scaler)

        if params.xia2.settings.multi_sweep_indexing is Auto:
            if (params.xia2.settings.small_molecule is True
                    and "dials" == params.xia2.settings.indexer):
                PhilIndex.update("xia2.settings.multi_sweep_indexing=True")
            else:
                PhilIndex.update("xia2.settings.multi_sweep_indexing=False")
        if (params.xia2.settings.multi_sweep_indexing is True
                and params.xia2.settings.multiprocessing.mode == "parallel"):
            Chatter.write(
                "Multi sweep indexing disabled:\nMSI is not available for parallel processing."
            )
            PhilIndex.update("xia2.settings.multi_sweep_indexing=False")

        input_json = params.xia2.settings.input.json
        if input_json is not None and len(input_json):
            for json_file in input_json:
                assert os.path.isfile(json_file)
                load_experiments(json_file)

        reference_geometry = params.xia2.settings.input.reference_geometry
        if reference_geometry is not None and len(reference_geometry) > 0:
            reference_geometries = "\n".join([
                "xia2.settings.input.reference_geometry=%s" %
                os.path.abspath(g)
                for g in params.xia2.settings.input.reference_geometry
            ])
            Debug.write(reference_geometries)
            PhilIndex.update(reference_geometries)
            Debug.write("xia2.settings.trust_beam_centre=true")
            PhilIndex.update("xia2.settings.trust_beam_centre=true")
            params = PhilIndex.get_python_object()

        params = PhilIndex.get_python_object()
        if params.xia2.settings.input.xinfo is not None:
            xinfo_file = os.path.abspath(params.xia2.settings.input.xinfo)
            PhilIndex.update("xia2.settings.input.xinfo=%s" % xinfo_file)
            params = PhilIndex.get_python_object()
            self.set_xinfo(xinfo_file)

            # issue #55 if not set ATOM in xinfo but anomalous=true or atom= set
            # on commandline, set here, should be idempotent

            if params.xia2.settings.input.anomalous is True:
                crystals = self._xinfo.get_crystals()
                for xname in crystals:
                    xtal = crystals[xname]
                    Debug.write("Setting anomalous for crystal %s" % xname)
                    xtal.set_anomalous(True)
        else:
            xinfo_file = "%s/automatic.xinfo" % os.path.abspath(os.curdir)
            PhilIndex.update("xia2.settings.input.xinfo=%s" % xinfo_file)
            params = PhilIndex.get_python_object()

        if params.dials.find_spots.phil_file is not None:
            PhilIndex.update(
                "dials.find_spots.phil_file=%s" %
                os.path.abspath(params.dials.find_spots.phil_file))
        if params.dials.index.phil_file is not None:
            PhilIndex.update("dials.index.phil_file=%s" %
                             os.path.abspath(params.dials.index.phil_file))
        if params.dials.refine.phil_file is not None:
            PhilIndex.update("dials.refine.phil_file=%s" %
                             os.path.abspath(params.dials.refine.phil_file))
        if params.dials.integrate.phil_file is not None:
            PhilIndex.update("dials.integrate.phil_file=%s" %
                             os.path.abspath(params.dials.integrate.phil_file))
        if params.xds.index.xparm is not None:
            Flags.set_xparm(params.xds.index.xparm)
        if params.xds.index.xparm_ub is not None:
            Flags.set_xparm_ub(params.xds.index.xparm_ub)

        if params.xia2.settings.scale.freer_file is not None:
            freer_file = os.path.abspath(params.xia2.settings.scale.freer_file)
            if not os.path.exists(freer_file):
                raise RuntimeError("%s does not exist" % freer_file)
            from xia2.Modules.FindFreeFlag import FindFreeFlag

            column = FindFreeFlag(freer_file)
            Debug.write("FreeR_flag column in %s found: %s" %
                        (freer_file, column))
            PhilIndex.update("xia2.settings.scale.freer_file=%s" % freer_file)

        if params.xia2.settings.scale.reference_reflection_file is not None:
            reference_reflection_file = os.path.abspath(
                params.xia2.settings.scale.reference_reflection_file)
            if not os.path.exists(reference_reflection_file):
                raise RuntimeError("%s does not exist" %
                                   reference_reflection_file)
            PhilIndex.update(
                "xia2.settings.scale.reference_reflection_file=%s" %
                reference_reflection_file)

        params = PhilIndex.get_python_object()

        datasets = unroll_datasets(PhilIndex.params.xia2.settings.input.image)

        for dataset in datasets:

            start_end = None

            # here we only care about ':' which are later than C:\
            if ":" in dataset[3:]:
                tokens = dataset.split(":")
                # cope with windows drives i.e. C:\data\blah\thing_0001.cbf:1:100
                if len(tokens[0]) == 1:
                    tokens = ["%s:%s" % (tokens[0], tokens[1])] + tokens[2:]
                if len(tokens) != 3:
                    raise RuntimeError("/path/to/image_0001.cbf:start:end")

                dataset = tokens[0]
                start_end = int(tokens[1]), int(tokens[2])

            from xia2.Applications.xia2setup import is_hd5f_name

            if os.path.exists(os.path.abspath(dataset)):
                dataset = os.path.abspath(dataset)
            else:
                directories = [os.getcwd()] + self._argv[1:]
                found = False
                for d in directories:
                    if os.path.exists(os.path.join(d, dataset)):
                        dataset = os.path.join(d, dataset)
                        found = True
                        break
                if not found:
                    raise Sorry("Could not find %s in %s" %
                                (dataset, " ".join(directories)))

            if is_hd5f_name(dataset):
                self._hdf5_master_files.append(dataset)
                if start_end:
                    Debug.write("Image range: %d %d" % start_end)
                    if dataset not in self._default_start_end:
                        self._default_start_end[dataset] = []
                    self._default_start_end[dataset].append(start_end)
                else:
                    Debug.write("No image range specified")

            else:
                template, directory = image2template_directory(
                    os.path.abspath(dataset))

                self._default_template.append(os.path.join(
                    directory, template))
                self._default_directory.append(directory)

                Debug.write("Interpreted from image %s:" % dataset)
                Debug.write("Template %s" % template)
                Debug.write("Directory %s" % directory)

                if start_end:
                    Debug.write("Image range: %d %d" % start_end)
                    key = os.path.join(directory, template)
                    if key not in self._default_start_end:
                        self._default_start_end[key] = []
                    self._default_start_end[key].append(start_end)
                else:
                    Debug.write("No image range specified")

        # finally, check that all arguments were read and raise an exception
        # if any of them were nonsense.

        with open("xia2-working.phil", "wb") as f:
            f.write(PhilIndex.working_phil.as_str())
            f.write(
                os.linesep
            )  # temporarily required for https://github.com/dials/dials/issues/522
        with open("xia2-diff.phil", "wb") as f:
            f.write(PhilIndex.get_diff().as_str())
            f.write(
                os.linesep
            )  # temporarily required for https://github.com/dials/dials/issues/522

        Debug.write("\nDifference PHIL:")
        Debug.write(PhilIndex.get_diff().as_str(), strip=False)

        Debug.write("Working PHIL:")
        Debug.write(PhilIndex.working_phil.as_str(), strip=False)

        nonsense = "Unknown command-line options:"
        was_nonsense = False

        for j, argv in enumerate(self._argv):
            if j == 0:
                continue
            if argv[0] != "-" and "=" not in argv:
                continue
            if j not in self._understood:
                nonsense += " %s" % argv
                was_nonsense = True

        if was_nonsense:
            raise RuntimeError(nonsense)
Exemplo n.º 2
0
def multi_crystal_analysis(stop_after=None):
  '''Actually process something...'''

  assert os.path.exists('xia2.json')
  from xia2.Schema.XProject import XProject
  xinfo = XProject.from_json(filename='xia2.json')

  crystals = xinfo.get_crystals()
  for crystal_id, crystal in crystals.iteritems():
    cwd = os.path.abspath(os.curdir)
    working_directory = Environment.generate_directory(
      [crystal.get_name(), 'analysis'])
    os.chdir(working_directory)

    from xia2.Wrappers.CCP4.Blend import Blend

    from xia2.lib.bits import auto_logfiler
    hand_blender = Blend()
    hand_blender.set_working_directory(working_directory)
    auto_logfiler(hand_blender)
    Citations.cite('blend')

    scaler = crystal._get_scaler()

    #epoch_to_si = {}
    epoch_to_batches = {}
    epoch_to_integrated_intensities = {}
    epoch_to_sweep_name = {}

    try:
      epochs = scaler._sweep_information.keys()
      for epoch in epochs:
        si = scaler._sweep_information[epoch]
        epoch_to_batches[epoch] = si['batches']
        epoch_to_integrated_intensities[epoch] = si['corrected_intensities']
        epoch_to_sweep_name[epoch] = si['sname']
    except AttributeError, e:
      epochs = scaler._sweep_handler.get_epochs()
      for epoch in epochs:
        si = scaler._sweep_handler.get_sweep_information(epoch)
        epoch_to_batches[epoch] = si.get_batches()
        epoch_to_integrated_intensities[epoch] = si.get_reflections()
        epoch_to_sweep_name[epoch] = si.get_sweep_name()

    unmerged_mtz = scaler.get_scaled_reflections('mtz_unmerged').values()[0]
    from iotbx.reflection_file_reader import any_reflection_file
    reader = any_reflection_file(unmerged_mtz)

    intensities = None
    batches = None
    assert reader.file_type() == 'ccp4_mtz'
    arrays = reader.as_miller_arrays(merge_equivalents=False)
    for ma in arrays:
      if ma.info().labels == ['BATCH']:
        batches = ma
      elif ma.info().labels == ['I', 'SIGI']:
        intensities = ma
      elif ma.info().labels == ['I(+)', 'SIGI(+)', 'I(-)', 'SIGI(-)']:
        intensities = ma

    from xia2.Handlers.Environment import which
    Rscript_binary = which('Rscript', debug=False)
    if Rscript_binary is None:
      Chatter.write('Skipping BLEND analysis: Rscript not available')
    else:
      for epoch in epochs:
        hand_blender.add_hklin(epoch_to_integrated_intensities[epoch],
                               label=epoch_to_sweep_name[epoch])
      hand_blender.analysis()
      Chatter.write("Dendrogram saved to: %s" %hand_blender.get_dendrogram_file())
      analysis = hand_blender.get_analysis()
      summary = hand_blender.get_summary()
      clusters = hand_blender.get_clusters()

      linkage_matrix = hand_blender.get_linkage_matrix()
      ddict = hand_blender.plot_dendrogram()

      rows = []
      headers = ['Cluster', 'Datasets', 'Multiplicity', 'Completeness', 'LCV', 'aLCV']
      completeness = flex.double()
      for i, cluster in clusters.iteritems():
        sel_cluster = flex.bool(batches.size(), False)
        for j in cluster['dataset_ids']:
          batch_start, batch_end = epoch_to_batches[epochs[j-1]]
          sel_cluster |= (
            (batches.data() >= batch_start) & (batches.data() <= batch_end))
        intensities_cluster = intensities.select(sel_cluster)
        merging = intensities_cluster.merge_equivalents()
        merged_intensities = merging.array()
        multiplicities = merging.redundancies()
        completeness.append(merged_intensities.completeness())
        dataset_ids = cluster['dataset_ids']

        rows.append(
          ['%i' %i, ' '.join(['%i'] * len(dataset_ids)) %tuple(dataset_ids),
           '%.1f' %flex.mean(multiplicities.data().as_double()),
           '%.2f' %completeness[-1],
           '%.2f' %cluster['lcv'], '%.2f' %cluster['alcv']])

      # sort table by completeness
      perm = flex.sort_permutation(completeness)
      rows = [rows[i] for i in perm]

      print
      print 'Unit cell clustering summary:'
      print tabulate(rows, headers, tablefmt='rst')
      print

      blend_html = tabulate(rows, headers, tablefmt='html').replace(
        '<table>', '<table class="table table-hover table-condensed">').replace(
    '<td>', '<td style="text-align: right;">')
Exemplo n.º 3
0
def multi_crystal_analysis(stop_after=None):
    '''Actually process something...'''

    assert os.path.exists('xia2.json')
    from xia2.Schema.XProject import XProject
    xinfo = XProject.from_json(filename='xia2.json')

    crystals = xinfo.get_crystals()
    for crystal_id, crystal in crystals.iteritems():
        cwd = os.path.abspath(os.curdir)
        working_directory = Environment.generate_directory(
            [crystal.get_name(), 'analysis'])
        os.chdir(working_directory)

        scaler = crystal._get_scaler()

        #epoch_to_si = {}
        epoch_to_batches = {}
        epoch_to_integrated_intensities = {}
        epoch_to_sweep_name = {}
        epoch_to_experiments_filename = {}
        epoch_to_experiments = {}
        sweep_name_to_epoch = {}
        epoch_to_first_image = {}

        from dxtbx.serialize import load
        try:
            epochs = scaler._sweep_information.keys()
            for epoch in epochs:
                si = scaler._sweep_information[epoch]
                epoch_to_batches[epoch] = si['batches']
                epoch_to_integrated_intensities[epoch] = si[
                    'corrected_intensities']
                epoch_to_sweep_name[epoch] = si['sname']
                sweep_name_to_epoch[si['name']] = epoch
                intgr = si['integrater']
                epoch_to_experiments_filename[
                    epoch] = intgr.get_integrated_experiments()
                epoch_to_experiments[epoch] = load.experiment_list(
                    intgr.get_integrated_experiments())

        except AttributeError:
            epochs = scaler._sweep_handler.get_epochs()
            for epoch in epochs:
                si = scaler._sweep_handler.get_sweep_information(epoch)
                epoch_to_batches[epoch] = si.get_batches()
                epoch_to_integrated_intensities[epoch] = si.get_reflections()
                epoch_to_sweep_name[epoch] = si.get_sweep_name()
                sweep_name_to_epoch[si.get_sweep_name()] = epoch
                intgr = si.get_integrater()
                epoch_to_experiments_filename[
                    epoch] = intgr.get_integrated_experiments()
                epoch_to_experiments[epoch] = load.experiment_list(
                    intgr.get_integrated_experiments())

        from xia2.Wrappers.Dials.StereographicProjection import StereographicProjection
        sp_json_files = {}
        for hkl in ((1, 0, 0), (0, 1, 0), (0, 0, 1)):
            sp = StereographicProjection()
            auto_logfiler(sp)
            sp.set_working_directory(working_directory)
            for experiments in epoch_to_experiments_filename.values():
                sp.add_experiments(experiments)
            sp.set_hkl(hkl)
            sp.run()
            sp_json_files[hkl] = sp.get_json_filename()

        unmerged_mtz = scaler.get_scaled_reflections(
            'mtz_unmerged').values()[0]
        from iotbx.reflection_file_reader import any_reflection_file
        reader = any_reflection_file(unmerged_mtz)

        from xia2.Wrappers.XIA.PlotMultiplicity import PlotMultiplicity
        mult_json_files = {}
        for axis in ('h', 'k', 'l'):
            pm = PlotMultiplicity()
            auto_logfiler(pm)
            pm.set_working_directory(working_directory)
            pm.set_mtz_filename(unmerged_mtz)
            pm.set_slice_axis(axis)
            pm.set_show_missing(True)
            pm.run()
            mult_json_files[axis] = pm.get_json_filename()

        intensities = None
        batches = None
        assert reader.file_type() == 'ccp4_mtz'
        arrays = reader.as_miller_arrays(merge_equivalents=False)
        for ma in arrays:
            if ma.info().labels == ['BATCH']:
                batches = ma
            elif ma.info().labels == ['I', 'SIGI']:
                intensities = ma
            elif ma.info().labels == ['I(+)', 'SIGI(+)', 'I(-)', 'SIGI(-)']:
                intensities = ma

        from xia2.Wrappers.CCP4.Blend import Blend
        hand_blender = Blend()
        hand_blender.set_working_directory(working_directory)
        auto_logfiler(hand_blender)
        Citations.cite('blend')

        from xia2.Handlers.Environment import which
        Rscript_binary = which('Rscript', debug=False)
        if Rscript_binary is None:
            Chatter.write('Skipping BLEND analysis: Rscript not available')
        else:
            for epoch in epochs:
                hand_blender.add_hklin(epoch_to_integrated_intensities[epoch],
                                       label=epoch_to_sweep_name[epoch])
            hand_blender.analysis()
            Chatter.write("Dendrogram saved to: %s" %
                          hand_blender.get_dendrogram_file())
            analysis = hand_blender.get_analysis()
            summary = hand_blender.get_summary()
            clusters = hand_blender.get_clusters()

            ddict = hand_blender.plot_dendrogram()

            phil_files_dir = 'phil_files'
            if not os.path.exists(phil_files_dir):
                os.makedirs(phil_files_dir)

            rows = []
            headers = [
                'Cluster', 'Datasets', 'Multiplicity', 'Completeness', 'LCV',
                'aLCV', 'Average unit cell'
            ]
            completeness = flex.double()
            average_unit_cell_params = []
            for i, cluster in clusters.iteritems():
                print i
                sel_cluster = flex.bool(batches.size(), False)
                cluster_uc_params = [flex.double() for k in range(6)]
                for j in cluster['dataset_ids']:
                    epoch = epochs[j - 1]
                    batch_start, batch_end = epoch_to_batches[epoch]
                    sel_cluster |= ((batches.data() >= batch_start) &
                                    (batches.data() <= batch_end))
                    expts = epoch_to_experiments.get(epoch)
                    assert expts is not None, (epoch)
                    assert len(expts) == 1, len(expts)
                    expt = expts[0]
                    uc_params = expt.crystal.get_unit_cell().parameters()
                    for k in range(6):
                        cluster_uc_params[k].append(uc_params[k])
                intensities_cluster = intensities.select(sel_cluster)
                merging = intensities_cluster.merge_equivalents()
                merged_intensities = merging.array()
                multiplicities = merging.redundancies()
                completeness.append(merged_intensities.completeness())
                average_unit_cell_params.append(
                    tuple(flex.mean(p) for p in cluster_uc_params))
                dataset_ids = cluster['dataset_ids']

                assert min(dataset_ids) > 0
                with open(
                        os.path.join(phil_files_dir,
                                     'blend_cluster_%i_images.phil' % i),
                        'wb') as f:
                    sweep_names = [
                        hand_blender._labels[dataset_id - 1]
                        for dataset_id in dataset_ids
                    ]
                    for sweep_name in sweep_names:
                        expts = epoch_to_experiments.get(
                            sweep_name_to_epoch.get(sweep_name))
                        assert expts is not None, (
                            sweep_name, sweep_name_to_epoch.get(sweep_name))
                        assert len(expts) == 1, len(expts)
                        expt = expts[0]
                        print >> f, 'xia2.settings.input.image = %s' % expt.imageset.get_path(
                            0)

                rows.append([
                    '%i' % i,
                    ' '.join(['%i'] * len(dataset_ids)) % tuple(dataset_ids),
                    '%.1f' % flex.mean(multiplicities.data().as_double()),
                    '%.2f' % completeness[-1],
                    '%.2f' % cluster['lcv'],
                    '%.2f' % cluster['alcv'],
                    '%g %g %g %g %g %g' % average_unit_cell_params[-1]
                ])

            # sort table by completeness
            perm = flex.sort_permutation(completeness)
            rows = [rows[i] for i in perm]

            print
            print 'Unit cell clustering summary:'
            print tabulate(rows, headers, tablefmt='rst')
            print

            blend_html = tabulate(rows, headers, tablefmt='html').replace(
                '<table>',
                '<table class="table table-hover table-condensed">').replace(
                    '<td>', '<td style="text-align: right;">')

    # XXX what about multiple wavelengths?
    with open('batches.phil', 'wb') as f:
        try:
            for epoch, si in scaler._sweep_information.iteritems():
                print >> f, "batch {"
                print >> f, "  id=%s" % si['sname']
                print >> f, "  range=%i,%i" % tuple(si['batches'])
                print >> f, "}"
        except AttributeError:
            for epoch in scaler._sweep_handler.get_epochs():
                si = scaler._sweep_handler.get_sweep_information(epoch)
                print >> f, "batch {"
                print >> f, "  id=%s" % si.get_sweep_name()
                print >> f, "  range=%i,%i" % tuple(si.get_batches())
                print >> f, "}"

    from xia2.Wrappers.XIA.MultiCrystalAnalysis import MultiCrystalAnalysis
    mca = MultiCrystalAnalysis()
    auto_logfiler(mca, extra="MultiCrystalAnalysis")
    mca.add_command_line_args([
        scaler.get_scaled_reflections(format="sca_unmerged").values()[0],
        "unit_cell=%s %s %s %s %s %s" % tuple(scaler.get_scaler_cell()),
        "batches.phil"
    ])
    mca.set_working_directory(working_directory)
    mca.run()

    intensity_clusters = mca.get_clusters()
    rows = []
    headers = [
        'Cluster', 'Datasets', 'Multiplicity', 'Completeness', 'Height',
        'Average unit cell'
    ]
    completeness = flex.double()
    average_unit_cell_params = []
    for i, cluster in intensity_clusters.iteritems():
        sel_cluster = flex.bool(batches.size(), False)
        cluster_uc_params = [flex.double() for k in range(6)]
        for j in cluster['datasets']:
            epoch = epochs[j - 1]
            batch_start, batch_end = epoch_to_batches[epoch]
            sel_cluster |= ((batches.data() >= batch_start) &
                            (batches.data() <= batch_end))
            expts = epoch_to_experiments.get(epoch)
            assert expts is not None, (epoch)
            assert len(expts) == 1, len(expts)
            expt = expts[0]
            uc_params = expt.crystal.get_unit_cell().parameters()
            for k in range(6):
                cluster_uc_params[k].append(uc_params[k])
        intensities_cluster = intensities.select(sel_cluster)
        merging = intensities_cluster.merge_equivalents()
        merged_intensities = merging.array()
        multiplicities = merging.redundancies()
        completeness.append(merged_intensities.completeness())
        average_unit_cell_params.append(
            tuple(flex.mean(p) for p in cluster_uc_params))
        dataset_ids = cluster['datasets']

        rows.append([
            '%i' % int(i),
            ' '.join(['%i'] * len(dataset_ids)) % tuple(dataset_ids),
            '%.1f' % flex.mean(multiplicities.data().as_double()),
            '%.2f' % completeness[-1],
            '%.2f' % cluster['height'],
            '%g %g %g %g %g %g' % average_unit_cell_params[-1]
        ])

    # sort table by completeness
    perm = flex.sort_permutation(completeness)
    rows = [rows[i] for i in perm]

    print 'Intensity clustering summary:'
    print tabulate(rows, headers, tablefmt='rst')
    print

    intensity_clustering_html = tabulate(
        rows, headers, tablefmt='html').replace(
            '<table>',
            '<table class="table table-hover table-condensed">').replace(
                '<td>', '<td style="text-align: right;">')

    import json

    json_data = {}
    if ddict is not None:
        from xia2.Modules.MultiCrystalAnalysis import scipy_dendrogram_to_plotly_json
        json_data['blend_dendrogram'] = scipy_dendrogram_to_plotly_json(ddict)
    else:
        json_data['blend_dendrogram'] = {'data': [], 'layout': {}}

    json_data['intensity_clustering'] = mca.get_dict()
    del json_data['intensity_clustering']['clusters']

    for hkl in ((1, 0, 0), (0, 1, 0), (0, 0, 1)):
        with open(sp_json_files[hkl], 'rb') as f:
            d = json.load(f)
            d['layout'][
                'title'] = 'Stereographic projection (hkl=%i%i%i)' % hkl
            json_data['stereographic_projection_%s%s%s' % hkl] = d

    for axis in ('h', 'k', 'l'):
        with open(mult_json_files[axis], 'rb') as f:
            json_data['multiplicity_%s' % axis] = json.load(f)

    json_str = json.dumps(json_data, indent=2)

    javascript = ['var graphs = %s' % (json_str)]
    javascript.append(
        'Plotly.newPlot(blend_dendrogram, graphs.blend_dendrogram.data, graphs.blend_dendrogram.layout);'
    )
    javascript.append(
        'Plotly.newPlot(intensity_clustering, graphs.intensity_clustering.data, graphs.intensity_clustering.layout);'
    )
    for hkl in ((1, 0, 0), (0, 1, 0), (0, 0, 1)):
        javascript.append(
            'Plotly.newPlot(stereographic_projection_%(hkl)s, graphs.stereographic_projection_%(hkl)s.data, graphs.stereographic_projection_%(hkl)s.layout);'
            % ({
                'hkl': "%s%s%s" % hkl
            }))
    for axis in ('h', 'k', 'l'):
        javascript.append(
            'Plotly.newPlot(multiplicity_%(axis)s, graphs.multiplicity_%(axis)s.data, graphs.multiplicity_%(axis)s.layout);'
            % ({
                'axis': axis
            }))

    html_header = '''
<head>

<!-- Plotly.js -->
<script src="https://cdn.plot.ly/plotly-latest.min.js"></script>

<meta name="viewport" content="width=device-width, initial-scale=1" charset="UTF-8">
<link rel="stylesheet" href="http://maxcdn.bootstrapcdn.com/bootstrap/3.3.5/css/bootstrap.min.css"/>
<script src="https://ajax.googleapis.com/ajax/libs/jquery/1.11.3/jquery.min.js"></script>
<script src="http://maxcdn.bootstrapcdn.com/bootstrap/3.3.5/js/bootstrap.min.js"></script>
<style type="text/css">

body {
  /*font-family: Helmet, Freesans, Helvetica, Arial, sans-serif;*/
  margin: 8px;
  min-width: 240px;
  margin-left: 5%;
  margin-right: 5%;
}

.plot {
  float: left;
  width: 1200px;
  height: 800px;
  margin-bottom: 20px;
}

.square_plot {
  float: left;
  width: 800px;
  height: 800px;
  margin-bottom: 20px;
}

</style>

</head>

'''

    html_body = '''

<body>

<div class="page-header">
  <h1>Multi-crystal analysis report</h1>
</div>

<div class="panel-group">

  <div class="panel panel-default">
    <div class="panel-heading" data-toggle="collapse" href="#collapse_multiplicity">
      <h4 class="panel-title">
        <a>Multiplicity plots</a>
      </h4>
    </div>
    <div id="collapse_multiplicity" class="panel-collapse collapse">
      <div class="panel-body">
        <div class="col-xs-12 col-sm-12 col-md-12 square_plot" id="multiplicity_h"></div>
        <div class="col-xs-12 col-sm-12 col-md-12 square_plot" id="multiplicity_k"></div>
        <div class="col-xs-12 col-sm-12 col-md-12 square_plot" id="multiplicity_l"></div>
      </div>
    </div>
  </div>

  <div class="panel panel-default">
    <div class="panel-heading" data-toggle="collapse" href="#collapse_stereographic_projection">
      <h4 class="panel-title">
        <a>Stereographic projections</a>
      </h4>
    </div>
    <div id="collapse_stereographic_projection" class="panel-collapse collapse">
      <div class="panel-body">
        <div class="col-xs-12 col-sm-12 col-md-12 square_plot" id="stereographic_projection_100"></div>
        <div class="col-xs-12 col-sm-12 col-md-12 square_plot" id="stereographic_projection_010"></div>
        <div class="col-xs-12 col-sm-12 col-md-12 square_plot" id="stereographic_projection_001"></div>
      </div>
    </div>
  </div>

  <div class="panel panel-default">
    <div class="panel-heading" data-toggle="collapse" href="#collapse_cell">
      <h4 class="panel-title">
        <a>Unit cell clustering</a>
      </h4>
    </div>
    <div id="collapse_cell" class="panel-collapse collapse">
      <div class="panel-body">
        <div class="col-xs-12 col-sm-12 col-md-12 plot" id="blend_dendrogram"></div>
        <div class="table-responsive" style="width: 800px">
          %(blend_html)s
        </div>
      </div>
    </div>
  </div>

  <div class="panel panel-default">
    <div class="panel-heading" data-toggle="collapse" href="#collapse_intensity">
      <h4 class="panel-title">
        <a>Intensity clustering</a>
      </h4>
    </div>
    <div id="collapse_intensity" class="panel-collapse collapse">
      <div class="panel-body">
        <div class="col-xs-12 col-sm-12 col-md-12 plot" id="intensity_clustering" style="height:1000px"></div>
        <div class="table-responsive" style="width: 800px">
          %(intensity_clustering_html)s
        </div>
      </div>
    </div>
  </div>
</div>

<script>
%(script)s
</script>
</body>
    ''' % {
        'script': '\n'.join(javascript),
        'blend_html': blend_html,
        'intensity_clustering_html': intensity_clustering_html
    }

    html = '\n'.join([html_header, html_body])

    print "Writing html report to: %s" % 'multi-crystal-report.html'
    with open('multi-crystal-report.html', 'wb') as f:
        print >> f, html.encode('ascii', 'xmlcharrefreplace')

    write_citations()

    Environment.cleanup()

    return
Exemplo n.º 4
0
    def setup(self):
        '''Set everything up...'''

        # check arguments are all ascii

        Debug.write('Start parsing command line: ' + str(sys.argv))

        for token in sys.argv:
            try:
                token.encode('ascii')
            except UnicodeDecodeError:
                raise RuntimeError('non-ascii characters in input')

        self._argv = copy.deepcopy(sys.argv)

        replacements = {
            '-2d': 'pipeline=2d',
            '-2di': 'pipeline=2di',
            '-3d': 'pipeline=3d',
            '-3di': 'pipeline=3di',
            '-3dii': 'pipeline=3dii',
            '-3dd': 'pipeline=3dd',
            '-dials': 'pipeline=dials',
            '-quick': 'dials.fast_mode=true',
            '-failover': 'failover=true',
            '-small_molecule': 'small_molecule=true'
        }
        for k, v in replacements.iteritems():
            if k in self._argv:
                print "***\nCommand line option %s is deprecated.\nPlease use %s instead\n***" % (
                    k, v)
                self._argv[self._argv.index(k)] = v
        if '-atom' in self._argv:
            idx = self._argv.index('-atom')
            element = self._argv[idx + 1]
            self._argv[idx:idx + 2] = ['atom=%s' % element]
            print "***\nCommand line option -atom %s is deprecated.\nPlease use atom=%s instead\n***" % (
                element, element)

        # first of all try to interpret arguments as phil parameters/files

        from xia2.Handlers.Phil import master_phil
        from libtbx.phil import command_line
        cmd_line = command_line.argument_interpreter(master_phil=master_phil)
        working_phil, self._argv = cmd_line.process_and_fetch(
            args=self._argv, custom_processor="collect_remaining")

        PhilIndex.merge_phil(working_phil)
        try:
            params = PhilIndex.get_python_object()
        except RuntimeError as e:
            raise Sorry(e)

        # sanity check / interpret Auto in input
        from libtbx import Auto

        if params.xia2.settings.input.atom is None:
            if params.xia2.settings.input.anomalous is Auto:
                PhilIndex.update("xia2.settings.input.anomalous=false")
        else:
            if params.xia2.settings.input.anomalous == False:
                raise Sorry(
                    'Setting anomalous=false and atom type inconsistent')
            params.xia2.settings.input.anomalous = True
            PhilIndex.update("xia2.settings.input.anomalous=true")

        if params.xia2.settings.resolution.keep_all_reflections is Auto:
            if params.xia2.settings.small_molecule == True and \
               params.xia2.settings.resolution.d_min is None and \
               params.xia2.settings.resolution.d_max is None:
                PhilIndex.update(
                    "xia2.settings.resolution.keep_all_reflections=true")
            else:
                PhilIndex.update(
                    "xia2.settings.resolution.keep_all_reflections=false")

        if params.xia2.settings.small_molecule == True:
            Debug.write('Small molecule selected')
            if params.ccp4.pointless.chirality is None:
                PhilIndex.update("ccp4.pointless.chirality=nonchiral")
            params = PhilIndex.get_python_object()

        # pipeline options
        self._read_pipeline()

        Debug.write('Project: %s' % params.xia2.settings.project)
        Debug.write('Crystal: %s' % params.xia2.settings.crystal)

        # FIXME add some consistency checks in here e.g. that there are
        # images assigned, there is a lattice assigned if cell constants
        # are given and so on

        params = PhilIndex.get_python_object()
        mp_params = params.xia2.settings.multiprocessing
        from xia2.Handlers.Environment import get_number_cpus
        if mp_params.mode == 'parallel':
            if mp_params.type == 'qsub':
                if which('qsub') is None:
                    raise Sorry('qsub not available')
            if mp_params.njob is Auto:
                mp_params.njob = get_number_cpus()
                if mp_params.nproc is Auto:
                    mp_params.nproc = 1
            elif mp_params.nproc is Auto:
                mp_params.nproc = get_number_cpus()
        elif mp_params.mode == 'serial':
            if mp_params.type == 'qsub':
                if which('qsub') is None:
                    raise Sorry('qsub not available')
            if mp_params.njob is Auto:
                mp_params.njob = 1
            if mp_params.nproc is Auto:
                mp_params.nproc = get_number_cpus()

        PhilIndex.update("xia2.settings.multiprocessing.njob=%d" %
                         mp_params.njob)
        PhilIndex.update("xia2.settings.multiprocessing.nproc=%d" %
                         mp_params.nproc)
        params = PhilIndex.get_python_object()
        mp_params = params.xia2.settings.multiprocessing

        if params.xia2.settings.indexer is not None:
            add_preference("indexer", params.xia2.settings.indexer)
        if params.xia2.settings.refiner is not None:
            add_preference("refiner", params.xia2.settings.refiner)
        if params.xia2.settings.integrater is not None:
            add_preference("integrater", params.xia2.settings.integrater)
        if params.xia2.settings.scaler is not None:
            add_preference("scaler", params.xia2.settings.scaler)

        if params.xia2.settings.multi_sweep_indexing is Auto:
            if params.xia2.settings.small_molecule == True and 'dials' == params.xia2.settings.indexer:
                PhilIndex.update("xia2.settings.multi_sweep_indexing=True")
            else:
                PhilIndex.update("xia2.settings.multi_sweep_indexing=False")
        if params.xia2.settings.multi_sweep_indexing == True and \
           params.xia2.settings.multiprocessing.mode == 'parallel':
            Chatter.write(
                'Multi sweep indexing disabled:\nMSI is not available for parallel processing.'
            )
            PhilIndex.update("xia2.settings.multi_sweep_indexing=False")

        input_json = params.xia2.settings.input.json
        if (input_json is not None and len(input_json)):
            for json_file in input_json:
                assert os.path.isfile(json_file)
                load_datablock(json_file)

        reference_geometry = params.xia2.settings.input.reference_geometry
        if reference_geometry is not None and len(reference_geometry) > 0:
            reference_geometries = "\n".join([
                "xia2.settings.input.reference_geometry=%s" %
                os.path.abspath(g)
                for g in params.xia2.settings.input.reference_geometry
            ])
            Debug.write(reference_geometries)
            PhilIndex.update(reference_geometries)
            Debug.write("xia2.settings.trust_beam_centre=true")
            PhilIndex.update("xia2.settings.trust_beam_centre=true")
            params = PhilIndex.get_python_object()

        params = PhilIndex.get_python_object()
        if params.xia2.settings.input.xinfo is not None:
            xinfo_file = os.path.abspath(params.xia2.settings.input.xinfo)
            PhilIndex.update("xia2.settings.input.xinfo=%s" % xinfo_file)
            params = PhilIndex.get_python_object()
            self.set_xinfo(xinfo_file)

            # issue #55 if not set ATOM in xinfo but anomalous=true or atom= set
            # on commandline, set here, should be idempotent

            if params.xia2.settings.input.anomalous is True:
                crystals = self._xinfo.get_crystals()
                for xname in crystals:
                    xtal = crystals[xname]
                    Debug.write("Setting anomalous for crystal %s" % xname)
                    xtal.set_anomalous(True)
        else:
            xinfo_file = '%s/automatic.xinfo' % os.path.abspath(os.curdir)
            PhilIndex.update("xia2.settings.input.xinfo=%s" % xinfo_file)
            params = PhilIndex.get_python_object()

        if params.dials.find_spots.phil_file is not None:
            PhilIndex.update(
                "dials.find_spots.phil_file=%s" %
                os.path.abspath(params.dials.find_spots.phil_file))
        if params.dials.index.phil_file is not None:
            PhilIndex.update("dials.index.phil_file=%s" %
                             os.path.abspath(params.dials.index.phil_file))
        if params.dials.refine.phil_file is not None:
            PhilIndex.update("dials.refine.phil_file=%s" %
                             os.path.abspath(params.dials.refine.phil_file))
        if params.dials.integrate.phil_file is not None:
            PhilIndex.update("dials.integrate.phil_file=%s" %
                             os.path.abspath(params.dials.integrate.phil_file))
        if params.xds.index.xparm is not None:
            Flags.set_xparm(params.xds.index.xparm)
        if params.xds.index.xparm_ub is not None:
            Flags.set_xparm_ub(params.xds.index.xparm_ub)

        if params.xia2.settings.scale.freer_file is not None:
            freer_file = os.path.abspath(params.xia2.settings.scale.freer_file)
            if not os.path.exists(freer_file):
                raise RuntimeError('%s does not exist' % freer_file)
            from xia2.Modules.FindFreeFlag import FindFreeFlag
            column = FindFreeFlag(freer_file)
            Debug.write('FreeR_flag column in %s found: %s' % \
                        (freer_file, column))
            PhilIndex.update("xia2.settings.scale.freer_file=%s" % freer_file)

        if params.xia2.settings.scale.reference_reflection_file is not None:
            reference_reflection_file = os.path.abspath(
                params.xia2.settings.scale.reference_reflection_file)
            if not os.path.exists(reference_reflection_file):
                raise RuntimeError('%s does not exist' %
                                   reference_reflection_file)
            PhilIndex.update(
                "xia2.settings.scale.reference_reflection_file=%s" %
                reference_reflection_file)

        params = PhilIndex.get_python_object()

        datasets = unroll_datasets(PhilIndex.params.xia2.settings.input.image)

        for dataset in datasets:

            start_end = None

            if ':' in dataset:
                tokens = dataset.split(':')
                # cope with windows drives i.e. C:\data\blah\thing_0001.cbf:1:100
                if len(tokens[0]) == 1:
                    tokens = ['%s:%s' % (tokens[0], tokens[1])] + tokens[2:]
                if len(tokens) != 3:
                    raise RuntimeError('/path/to/image_0001.cbf:start:end')

                dataset = tokens[0]
                start_end = int(tokens[1]), int(tokens[2])

            from xia2.Applications.xia2setup import is_hd5f_name
            if os.path.exists(os.path.abspath(dataset)):
                dataset = os.path.abspath(dataset)
            else:
                directories = [os.getcwd()] + self._argv[1:]
                found = False
                for d in directories:
                    if os.path.exists(os.path.join(d, dataset)):
                        dataset = os.path.join(d, dataset)
                        found = True
                        break
                if not found:
                    raise Sorry('Cound not find %s in %s' % \
                                (dataset, ' '.join(directories)))

            if is_hd5f_name(dataset):
                self._hdf5_master_files.append(dataset)
                if start_end:
                    Debug.write('Image range: %d %d' % start_end)
                    if not dataset in self._default_start_end:
                        self._default_start_end[dataset] = []
                    self._default_start_end[dataset].append(start_end)
                else:
                    Debug.write('No image range specified')

            else:
                template, directory = image2template_directory(
                    os.path.abspath(dataset))

                self._default_template.append(os.path.join(
                    directory, template))
                self._default_directory.append(directory)

                Debug.write('Interpreted from image %s:' % dataset)
                Debug.write('Template %s' % template)
                Debug.write('Directory %s' % directory)

                if start_end:
                    Debug.write('Image range: %d %d' % start_end)
                    key = os.path.join(directory, template)
                    if not key in self._default_start_end:
                        self._default_start_end[key] = []
                    self._default_start_end[key].append(start_end)
                else:
                    Debug.write('No image range specified')

        # finally, check that all arguments were read and raise an exception
        # if any of them were nonsense.

        with open('xia2-working.phil', 'wb') as f:
            print >> f, PhilIndex.working_phil.as_str()
        with open('xia2-diff.phil', 'wb') as f:
            print >> f, PhilIndex.get_diff().as_str()

        Debug.write('\nDifference PHIL:')
        Debug.write(PhilIndex.get_diff().as_str(), strip=False)

        Debug.write('Working PHIL:')
        Debug.write(PhilIndex.working_phil.as_str(), strip=False)

        nonsense = 'Unknown command-line options:'
        was_nonsense = False

        for j, argv in enumerate(self._argv):
            if j == 0:
                continue
            if argv[0] != '-' and '=' not in argv:
                continue
            if not j in self._understood:
                nonsense += ' %s' % argv
                was_nonsense = True

        if was_nonsense:
            raise RuntimeError(nonsense)