source: sasview/src/sas/sascalc/dataloader/readers/cansas_reader_HDF5.py @ a14fa99

ESS_GUIESS_GUI_DocsESS_GUI_batch_fittingESS_GUI_bumps_abstractionESS_GUI_iss1116ESS_GUI_iss879ESS_GUI_iss959ESS_GUI_openclESS_GUI_orderingESS_GUI_sync_sascalccostrafo411magnetic_scattrelease-4.1.1release-4.1.2release-4.2.2ticket-1009ticket-1094-headlessticket-1242-2d-resolutionticket-1243ticket-1249ticket885unittest-saveload
Last change on this file since a14fa99 was 995f4eb, checked in by lewis, 8 years ago

Ensure CanSAS 2.0 files are closed after reading
  • Property mode set to 100644
File size: 20.1 KB
Line 
1"""
2    CanSAS 2D data reader for reading HDF5 formatted CanSAS files.
3"""
4
5import h5py
6import numpy as np
7import re
8import os
9import sys
10
11from sas.sascalc.dataloader.data_info import plottable_1D, plottable_2D, Data1D, Data2D, DataInfo, Process, Aperture
12from sas.sascalc.dataloader.data_info import Collimation, TransmissionSpectrum, Detector
13from sas.sascalc.dataloader.data_info import combine_data_info_with_plottable
14
15
16
17class Reader():
18    """
19    A class for reading in CanSAS v2.0 data files. The existing iteration opens Mantid generated HDF5 formatted files
20    with file extension .h5/.H5. Any number of data sets may be present within the file and any dimensionality of data
21    may be used. Currently 1D and 2D SAS data sets are supported, but future implementations will include 1D and 2D
22    SESANS data.
23
24    Any number of SASdata sets may be present in a SASentry and the data within can be either 1D I(Q) or 2D I(Qx, Qy).
25
26    :Dependencies:
27        The CanSAS HDF5 reader requires h5py => v2.5.0 or later.
28    """
29
30    ## CanSAS version
31    cansas_version = 2.0
32    ## Logged warnings or messages
33    logging = None
34    ## List of errors for the current data set
35    errors = None
36    ## Raw file contents to be processed
37    raw_data = None
38    ## Data info currently being read in
39    current_datainfo = None
40    ## SASdata set currently being read in
41    current_dataset = None
42    ## List of plottable1D objects that should be linked to the current_datainfo
43    data1d = None
44    ## List of plottable2D objects that should be linked to the current_datainfo
45    data2d = None
46    ## Data type name
47    type_name = "CanSAS 2.0"
48    ## Wildcards
49    type = ["CanSAS 2.0 HDF5 Files (*.h5)|*.h5"]
50    ## List of allowed extensions
51    ext = ['.h5', '.H5']
52    ## Flag to bypass extension check
53    allow_all = False
54    ## List of files to return
55    output = None
56
57    def read(self, filename):
58        """
59        This is the general read method that all SasView data_loaders must have.
60
61        :param filename: A path for an HDF5 formatted CanSAS 2D data file.
62        :return: List of Data1D/2D objects and/or a list of errors.
63        """
64        ## Reinitialize the class when loading a new data file to reset all class variables
65        self.reset_class_variables()
66        ## Check that the file exists
67        if os.path.isfile(filename):
68            basename = os.path.basename(filename)
69            _, extension = os.path.splitext(basename)
70            # If the file type is not allowed, return empty list
71            if extension in self.ext or self.allow_all:
72                ## Load the data file
73                self.raw_data = h5py.File(filename, 'r')
74                ## Read in all child elements of top level SASroot
75                self.read_children(self.raw_data, [])
76                ## Add the last data set to the list of outputs
77                self.add_data_set()
78                ## Close the data file
79                self.raw_data.close()
80        ## Return data set(s)
81        return self.output
82
83    def reset_class_variables(self):
84        """
85        Create the reader object and define initial states for class variables
86        """
87        self.current_datainfo = None
88        self.current_dataset = None
89        self.data1d = []
90        self.data2d = []
91        self.raw_data = None
92        self.errors = set()
93        self.logging = []
94        self.output = []
95        self.parent_class = u''
96        self.detector = Detector()
97        self.collimation = Collimation()
98        self.aperture = Aperture()
99        self.process = Process()
100        self.trans_spectrum = TransmissionSpectrum()
101
102    def read_children(self, data, parent_list):
103        """
104        A recursive method for stepping through the hierarchical data file.
105
106        :param data: h5py Group object of any kind
107        :param parent: h5py Group parent name
108        """
109
110        ## Loop through each element of the parent and process accordingly
111        for key in data.keys():
112            ## Get all information for the current key
113            value = data.get(key)
114            if value.attrs.get(u'canSAS_class') is not None:
115                class_name = value.attrs.get(u'canSAS_class')
116            else:
117                class_name = value.attrs.get(u'NX_class')
118            if class_name is not None:
119                class_prog = re.compile(class_name)
120            else:
121                class_prog = re.compile(value.name)
122
123            if isinstance(value, h5py.Group):
124                self.parent_class = class_name
125                parent_list.append(key)
126                ## If this is a new sasentry, store the current data sets and create a fresh Data1D/2D object
127                if class_prog.match(u'SASentry'):
128                    self.add_data_set(key)
129                elif class_prog.match(u'SASdata'):
130                    self._initialize_new_data_set(parent_list)
131                ## Recursion step to access data within the group
132                self.read_children(value, parent_list)
133                self.add_intermediate()
134                parent_list.remove(key)
135
136            elif isinstance(value, h5py.Dataset):
137                ## If this is a dataset, store the data appropriately
138                data_set = data[key][:]
139                unit = self._get_unit(value)
140
141                ## I and Q Data
142                if key == u'I':
143                    if type(self.current_dataset) is plottable_2D:
144                        self.current_dataset.data = data_set
145                        self.current_dataset.zaxis("Intensity", unit)
146                    else:
147                        self.current_dataset.y = data_set.flatten()
148                        self.current_dataset.yaxis("Intensity", unit)
149                    continue
150                elif key == u'Idev':
151                    if type(self.current_dataset) is plottable_2D:
152                        self.current_dataset.err_data = data_set.flatten()
153                    else:
154                        self.current_dataset.dy = data_set.flatten()
155                    continue
156                elif key == u'Q':
157                    self.current_dataset.xaxis("Q", unit)
158                    if type(self.current_dataset) is plottable_2D:
159                        self.current_dataset.q = data_set.flatten()
160                    else:
161                        self.current_dataset.x = data_set.flatten()
162                    continue
163                elif key == u'Qy':
164                    self.current_dataset.yaxis("Q_y", unit)
165                    self.current_dataset.qy_data = data_set.flatten()
166                    continue
167                elif key == u'Qydev':
168                    self.current_dataset.dqy_data = data_set.flatten()
169                    continue
170                elif key == u'Qx':
171                    self.current_dataset.xaxis("Q_x", unit)
172                    self.current_dataset.qx_data = data_set.flatten()
173                    continue
174                elif key == u'Qxdev':
175                    self.current_dataset.dqx_data = data_set.flatten()
176                    continue
177                elif key == u'Mask':
178                    self.current_dataset.mask = data_set.flatten()
179                    continue
180
181                for data_point in data_set:
182                    ## Top Level Meta Data
183                    if key == u'definition':
184                        self.current_datainfo.meta_data['reader'] = data_point
185                    elif key == u'run':
186                        self.current_datainfo.run.append(data_point)
187                    elif key == u'title':
188                        self.current_datainfo.title = data_point
189                    elif key == u'SASnote':
190                        self.current_datainfo.notes.append(data_point)
191
192                    ## Sample Information
193                    elif key == u'Title' and self.parent_class == u'SASsample': # CanSAS 2.0 format
194                        self.current_datainfo.sample.name = data_point
195                    elif key == u'name' and self.parent_class == u'SASsample': # NXcanSAS format
196                        self.current_datainfo.sample.name = data_point
197                    elif key == u'thickness' and self.parent_class == u'SASsample':
198                        self.current_datainfo.sample.thickness = data_point
199                    elif key == u'temperature' and self.parent_class == u'SASsample':
200                        self.current_datainfo.sample.temperature = data_point
201
202                    ## Instrumental Information
203                    elif key == u'name' and self.parent_class == u'SASinstrument':
204                        self.current_datainfo.instrument = data_point
205                    elif key == u'name' and self.parent_class == u'SASdetector':
206                        self.detector.name = data_point
207                    elif key == u'SDD' and self.parent_class == u'SASdetector':
208                        self.detector.distance = float(data_point)
209                        self.detector.distance_unit = unit
210                    elif key == u'SSD' and self.parent_class == u'SAScollimation':
211                        self.collimation.length = data_point
212                        self.collimation.length_unit = unit
213                    elif key == u'name' and self.parent_class == u'SAScollimation':
214                        self.collimation.name = data_point
215
216                    ## Process Information
217                    elif key == u'name' and self.parent_class == u'SASprocess':
218                        self.process.name = data_point
219                    elif key == u'Title' and self.parent_class == u'SASprocess': # CanSAS 2.0 format
220                        self.process.name = data_point
221                    elif key == u'name' and self.parent_class == u'SASprocess': # NXcanSAS format
222                        self.process.name = data_point
223                    elif key == u'description' and self.parent_class == u'SASprocess':
224                        self.process.description = data_point
225                    elif key == u'date' and self.parent_class == u'SASprocess':
226                        self.process.date = data_point
227                    elif self.parent_class == u'SASprocess':
228                        self.process.notes.append(data_point)
229
230                    ## Transmission Spectrum
231                    elif key == u'T' and self.parent_class == u'SAStransmission_spectrum':
232                        self.trans_spectrum.transmission.append(data_point)
233                    elif key == u'Tdev' and self.parent_class == u'SAStransmission_spectrum':
234                        self.trans_spectrum.transmission_deviation.append(data_point)
235                    elif key == u'lambda' and self.parent_class == u'SAStransmission_spectrum':
236                        self.trans_spectrum.wavelength.append(data_point)
237
238                    ## Other Information
239                    elif key == u'wavelength' and self.parent_class == u'SASdata':
240                        self.current_datainfo.source.wavelength = data_point
241                        self.current_datainfo.source.wavelength.unit = unit
242                    elif key == u'radiation' and self.parent_class == u'SASsource':
243                        self.current_datainfo.source.radiation = data_point
244                    elif key == u'transmission' and self.parent_class == u'SASdata':
245                        self.current_datainfo.sample.transmission = data_point
246
247                    ## Everything else goes in meta_data
248                    else:
249                        new_key = self._create_unique_key(self.current_datainfo.meta_data, key)
250                        self.current_datainfo.meta_data[new_key] = data_point
251
252            else:
253                ## I don't know if this reachable code
254                self.errors.add("ShouldNeverHappenException")
255
256    def add_intermediate(self):
257        """
258        This method stores any intermediate objects within the final data set after fully reading the set.
259
260        :param parent: The NXclass name for the h5py Group object that just finished being processed
261        """
262
263        if self.parent_class == u'SASprocess':
264            self.current_datainfo.process.append(self.process)
265            self.process = Process()
266        elif self.parent_class == u'SASdetector':
267            self.current_datainfo.detector.append(self.detector)
268            self.detector = Detector()
269        elif self.parent_class == u'SAStransmission_spectrum':
270            self.current_datainfo.trans_spectrum.append(self.trans_spectrum)
271            self.trans_spectrum = TransmissionSpectrum()
272        elif self.parent_class == u'SAScollimation':
273            self.current_datainfo.collimation.append(self.collimation)
274            self.collimation = Collimation()
275        elif self.parent_class == u'SASaperture':
276            self.collimation.aperture.append(self.aperture)
277            self.aperture = Aperture()
278        elif self.parent_class == u'SASdata':
279            if type(self.current_dataset) is plottable_2D:
280                self.data2d.append(self.current_dataset)
281            elif type(self.current_dataset) is plottable_1D:
282                self.data1d.append(self.current_dataset)
283
284    def final_data_cleanup(self):
285        """
286        Does some final cleanup and formatting on self.current_datainfo and all data1D and data2D objects and then
287        combines the data and info into Data1D and Data2D objects
288        """
289
290        ## Type cast data arrays to float64
291        if len(self.current_datainfo.trans_spectrum) > 0:
292            spectrum_list = []
293            for spectrum in self.current_datainfo.trans_spectrum:
294                spectrum.transmission = np.delete(spectrum.transmission, [0])
295                spectrum.transmission = spectrum.transmission.astype(np.float64)
296                spectrum.transmission_deviation = np.delete(spectrum.transmission_deviation, [0])
297                spectrum.transmission_deviation = spectrum.transmission_deviation.astype(np.float64)
298                spectrum.wavelength = np.delete(spectrum.wavelength, [0])
299                spectrum.wavelength = spectrum.wavelength.astype(np.float64)
300                if len(spectrum.transmission) > 0:
301                    spectrum_list.append(spectrum)
302            self.current_datainfo.trans_spectrum = spectrum_list
303
304        ## Append errors to dataset and reset class errors
305        self.current_datainfo.errors = self.errors
306        self.errors.clear()
307
308        ## Combine all plottables with datainfo and append each to output
309        ## Type cast data arrays to float64 and find min/max as appropriate
310        for dataset in self.data2d:
311            dataset.data = dataset.data.astype(np.float64)
312            dataset.err_data = dataset.err_data.astype(np.float64)
313            if dataset.qx_data is not None:
314                dataset.xmin = np.min(dataset.qx_data)
315                dataset.xmax = np.max(dataset.qx_data)
316                dataset.qx_data = dataset.qx_data.astype(np.float64)
317            if dataset.dqx_data is not None:
318                dataset.dqx_data = dataset.dqx_data.astype(np.float64)
319            if dataset.qy_data is not None:
320                dataset.ymin = np.min(dataset.qy_data)
321                dataset.ymax = np.max(dataset.qy_data)
322                dataset.qy_data = dataset.qy_data.astype(np.float64)
323            if dataset.dqy_data is not None:
324                dataset.dqy_data = dataset.dqy_data.astype(np.float64)
325            if dataset.q_data is not None:
326                dataset.q_data = dataset.q_data.astype(np.float64)
327            zeros = np.ones(dataset.data.size, dtype=bool)
328            try:
329                for i in range (0, dataset.mask.size - 1):
330                    zeros[i] = dataset.mask[i]
331            except:
332                self.errors.add(sys.exc_value)
333            dataset.mask = zeros
334            ## Calculate the actual Q matrix
335            try:
336                if dataset.q_data.size <= 1:
337                    dataset.q_data = np.sqrt(dataset.qx_data * dataset.qx_data + dataset.qy_data * dataset.qy_data)
338            except:
339                dataset.q_data = None
340
341            if dataset.data.ndim == 2:
342                (n_rows, n_cols) = dataset.data.shape
343                dataset.y_bins = dataset.qy_data[0::n_rows]
344                dataset.x_bins = dataset.qx_data[:n_cols]
345                dataset.data = dataset.data.flatten()
346
347            final_dataset = combine_data_info_with_plottable(dataset, self.current_datainfo)
348            self.output.append(final_dataset)
349
350        for dataset in self.data1d:
351            if dataset.x is not None:
352                dataset.x = dataset.x.astype(np.float64)
353                dataset.xmin = np.min(dataset.x)
354                dataset.xmax = np.max(dataset.x)
355            if dataset.y is not None:
356                dataset.y = dataset.y.astype(np.float64)
357                dataset.ymin = np.min(dataset.y)
358                dataset.ymax = np.max(dataset.y)
359            if dataset.dx is not None:
360                dataset.dx = dataset.dx.astype(np.float64)
361            if dataset.dxl is not None:
362                dataset.dxl = dataset.dxl.astype(np.float64)
363            if dataset.dxw is not None:
364                dataset.dxw = dataset.dxw.astype(np.float64)
365            if dataset.dy is not None:
366                dataset.dy = dataset.dy.astype(np.float64)
367            final_dataset = combine_data_info_with_plottable(dataset, self.current_datainfo)
368            self.output.append(final_dataset)
369
370    def add_data_set(self, key=""):
371        """
372        Adds the current_dataset to the list of outputs after preforming final processing on the data and then calls a
373        private method to generate a new data set.
374
375        :param key: NeXus group name for current tree level
376        """
377
378        if self.current_datainfo and self.current_dataset:
379            self.final_data_cleanup()
380        self.data1d = []
381        self.data2d = []
382        self.current_datainfo = DataInfo()
383
384
385    def _initialize_new_data_set(self, parent_list = None):
386        """
387        A private class method to generate a new 1D or 2D data object based on the type of data within the set.
388        Outside methods should call add_data_set() to be sure any existing data is stored properly.
389
390        :param parent_list: List of names of parent elements
391        """
392
393        if parent_list is None:
394            parent_list = []
395        if self._find_intermediate(parent_list, "Qx"):
396            self.current_dataset = plottable_2D()
397        else:
398            x = np.array(0)
399            y = np.array(0)
400            self.current_dataset = plottable_1D(x, y)
401        self.current_datainfo.filename = self.raw_data.filename
402
403    def _find_intermediate(self, parent_list, basename=""):
404        """
405        A private class used to find an entry by either using a direct key or knowing the approximate basename.
406
407        :param parent_list: List of parents to the current level in the HDF5 file
408        :param basename: Approximate name of an entry to search for
409        :return:
410        """
411
412        entry = False
413        key_prog = re.compile(basename)
414        top = self.raw_data
415        for parent in parent_list:
416            top = top.get(parent)
417        for key in top.keys():
418            if (key_prog.match(key)):
419                entry = True
420                break
421        return entry
422
423    def _create_unique_key(self, dictionary, name, numb=0):
424        """
425        Create a unique key value for any dictionary to prevent overwriting
426        Recurses until a unique key value is found.
427
428        :param dictionary: A dictionary with any number of entries
429        :param name: The index of the item to be added to dictionary
430        :param numb: The number to be appended to the name, starts at 0
431        :return: The new name for the dictionary entry
432        """
433        if dictionary.get(name) is not None:
434            numb += 1
435            name = name.split("_")[0]
436            name += "_{0}".format(numb)
437            name = self._create_unique_key(dictionary, name, numb)
438        return name
439
440    def _get_unit(self, value):
441        """
442        Find the unit for a particular value within the h5py dictionary
443
444        :param value: attribute dictionary for a particular value set
445        :return: unit for the value passed to the method
446        """
447        unit = value.attrs.get(u'units')
448        if unit == None:
449            unit = value.attrs.get(u'unit')
450        ## Convert the unit formats
451        if unit == "1/A":
452            unit = "A^{-1}"
453        elif unit == "1/cm":
454            unit = "cm^{-1}"
455        return unit
Note: See TracBrowser for help on using the repository browser.