| 1 | """ |
|---|
| 2 | NXcanSAS 1/2D data reader for writing HDF5 formatted NXcanSAS files. |
|---|
| 3 | """ |
|---|
| 4 | |
|---|
| 5 | import h5py |
|---|
| 6 | import numpy as np |
|---|
| 7 | import re |
|---|
| 8 | import os |
|---|
| 9 | |
|---|
| 10 | from sas.sascalc.dataloader.readers.cansas_reader_HDF5 import Reader |
|---|
| 11 | from sas.sascalc.dataloader.data_info import Data1D, Data2D |
|---|
| 12 | |
|---|
| 13 | class NXcanSASWriter(Reader): |
|---|
| 14 | """ |
|---|
| 15 | A class for writing in NXcanSAS data files. Any number of data sets may be |
|---|
| 16 | written to the file. Currently 1D and 2D SAS data sets are supported |
|---|
| 17 | |
|---|
| 18 | NXcanSAS spec: http://download.nexusformat.org/sphinx/classes/contributed_definitions/NXcanSAS.html |
|---|
| 19 | |
|---|
| 20 | :Dependencies: |
|---|
| 21 | The NXcanSAS writer requires h5py => v2.5.0 or later. |
|---|
| 22 | """ |
|---|
| 23 | |
|---|
| 24 | def write(self, dataset, filename): |
|---|
| 25 | """ |
|---|
| 26 | Write an array of Data1d or Data2D objects to an NXcanSAS file, as |
|---|
| 27 | one SASEntry with multiple SASData elements. The metadata of the first |
|---|
| 28 | elememt in the array will be written as the SASentry metadata |
|---|
| 29 | (detector, instrument, sample, etc). |
|---|
| 30 | |
|---|
| 31 | :param dataset: A list of Data1D or Data2D objects to write |
|---|
| 32 | :param filename: Where to write the NXcanSAS file |
|---|
| 33 | """ |
|---|
| 34 | |
|---|
| 35 | def _h5_string(string): |
|---|
| 36 | """ |
|---|
| 37 | Convert a string to a numpy string in a numpy array. This way it is |
|---|
| 38 | written to the HDF5 file as a fixed length ASCII string and is |
|---|
| 39 | compatible with the Reader read() method. |
|---|
| 40 | """ |
|---|
| 41 | if isinstance(string, np.ndarray): |
|---|
| 42 | return string |
|---|
| 43 | elif not isinstance(string, str): |
|---|
| 44 | string = str(string) |
|---|
| 45 | |
|---|
| 46 | return np.array([np.string_(string)]) |
|---|
| 47 | |
|---|
| 48 | def _write_h5_string(entry, value, key): |
|---|
| 49 | entry[key] = _h5_string(value) |
|---|
| 50 | |
|---|
| 51 | def _h5_float(x): |
|---|
| 52 | if not (isinstance(x, list)): |
|---|
| 53 | x = [x] |
|---|
| 54 | return np.array(x, dtype=np.float32) |
|---|
| 55 | |
|---|
| 56 | def _write_h5_float(entry, value, key): |
|---|
| 57 | entry.create_dataset(key, data=_h5_float(value)) |
|---|
| 58 | |
|---|
| 59 | def _write_h5_vector(entry, vector, names=['x_position', 'y_position'], |
|---|
| 60 | units=None, write_fn=_write_h5_string): |
|---|
| 61 | """ |
|---|
| 62 | Write a vector to an h5 entry |
|---|
| 63 | |
|---|
| 64 | :param entry: The H5Py entry to write to |
|---|
| 65 | :param vector: The Vector to write |
|---|
| 66 | :param names: What to call the x,y and z components of the vector |
|---|
| 67 | when writing to the H5Py entry |
|---|
| 68 | :param units: The units of the vector (optional) |
|---|
| 69 | :param write_fn: A function to convert the value to the required |
|---|
| 70 | format and write it to the H5Py entry, of the form |
|---|
| 71 | f(entry, value, name) (optional) |
|---|
| 72 | """ |
|---|
| 73 | if len(names) < 2: |
|---|
| 74 | raise ValueError("Length of names must be >= 2.") |
|---|
| 75 | |
|---|
| 76 | if vector.x is not None: |
|---|
| 77 | write_fn(entry, vector.x, names[0]) |
|---|
| 78 | if units is not None: |
|---|
| 79 | entry[names[0]].attrs['units'] = units |
|---|
| 80 | if vector.y is not None: |
|---|
| 81 | write_fn(entry, vector.y, names[1]) |
|---|
| 82 | if units is not None: |
|---|
| 83 | entry[names[1]].attrs['units'] = units |
|---|
| 84 | if len(names) == 3 and vector.z is not None: |
|---|
| 85 | write_fn(entry, vector.z, names[2]) |
|---|
| 86 | if units is not None: |
|---|
| 87 | entry[names[2]].attrs['units'] = units |
|---|
| 88 | |
|---|
| 89 | valid_data = all([isinstance(d, (Data1D, Data2D)) for d in dataset]) |
|---|
| 90 | if not valid_data: |
|---|
| 91 | raise ValueError("All entries of dataset must be Data1D or Data2D" |
|---|
| 92 | "objects") |
|---|
| 93 | |
|---|
| 94 | # Get run name and number from first Data object |
|---|
| 95 | data_info = dataset[0] |
|---|
| 96 | run_number = '' |
|---|
| 97 | run_name = '' |
|---|
| 98 | if len(data_info.run) > 0: |
|---|
| 99 | run_number = data_info.run[0] |
|---|
| 100 | if len(data_info.run_name) > 0: |
|---|
| 101 | run_name = data_info.run_name[run_number] |
|---|
| 102 | |
|---|
| 103 | f = h5py.File(filename, 'w') |
|---|
| 104 | sasentry = f.create_group('sasentry01') |
|---|
| 105 | sasentry['definition'] = _h5_string('NXcanSAS') |
|---|
| 106 | sasentry['run'] = _h5_string(run_number) |
|---|
| 107 | sasentry['run'].attrs['name'] = run_name |
|---|
| 108 | sasentry['title'] = _h5_string(data_info.title) |
|---|
| 109 | sasentry.attrs['canSAS_class'] = 'SASentry' |
|---|
| 110 | sasentry.attrs['version'] = '1.0' |
|---|
| 111 | |
|---|
| 112 | for i, data_obj in enumerate(dataset): |
|---|
| 113 | data_entry = sasentry.create_group("sasdata{0:0=2d}".format(i+1)) |
|---|
| 114 | data_entry.attrs['canSAS_class'] = 'SASdata' |
|---|
| 115 | if isinstance(data_obj, Data1D): |
|---|
| 116 | self._write_1d_data(data_obj, data_entry) |
|---|
| 117 | elif isinstance(data_obj, Data2D): |
|---|
| 118 | self._write_2d_data(data_obj, data_entry) |
|---|
| 119 | |
|---|
| 120 | data_info = dataset[0] |
|---|
| 121 | # Sample metadata |
|---|
| 122 | sample_entry = sasentry.create_group('sassample') |
|---|
| 123 | sample_entry.attrs['canSAS_class'] = 'SASsample' |
|---|
| 124 | sample_entry['ID'] = _h5_string(data_info.sample.name) |
|---|
| 125 | sample_attrs = ['thickness', 'temperature', 'transmission'] |
|---|
| 126 | for key in sample_attrs: |
|---|
| 127 | if getattr(data_info.sample, key) is not None: |
|---|
| 128 | sample_entry.create_dataset(key, |
|---|
| 129 | data=_h5_float(getattr(data_info.sample, key))) |
|---|
| 130 | _write_h5_vector(sample_entry, data_info.sample.position) |
|---|
| 131 | # NXcanSAS doesn't save information about pitch, only roll |
|---|
| 132 | # and yaw. The _write_h5_vector method writes vector.y, but we |
|---|
| 133 | # need to write vector.z for yaw |
|---|
| 134 | data_info.sample.orientation.y = data_info.sample.orientation.z |
|---|
| 135 | _write_h5_vector(sample_entry, data_info.sample.orientation, |
|---|
| 136 | names=['polar_angle', 'azimuthal_angle']) |
|---|
| 137 | if data_info.sample.details is not None\ |
|---|
| 138 | and data_info.sample.details != []: |
|---|
| 139 | details = None |
|---|
| 140 | if len(data_info.sample.details) > 1: |
|---|
| 141 | details = [np.string_(d) for d in data_info.sample.details] |
|---|
| 142 | details = np.array(details) |
|---|
| 143 | elif data_info.sample.details != []: |
|---|
| 144 | details = _h5_string(data_info.sample.details[0]) |
|---|
| 145 | if details is not None: |
|---|
| 146 | sample_entry.create_dataset('details', data=details) |
|---|
| 147 | |
|---|
| 148 | # Instrument metadata |
|---|
| 149 | instrument_entry = sasentry.create_group('sasinstrument') |
|---|
| 150 | instrument_entry.attrs['canSAS_class'] = 'SASinstrument' |
|---|
| 151 | instrument_entry['name'] = _h5_string(data_info.instrument) |
|---|
| 152 | |
|---|
| 153 | # Source metadata |
|---|
| 154 | source_entry = instrument_entry.create_group('sassource') |
|---|
| 155 | source_entry.attrs['canSAS_class'] = 'SASsource' |
|---|
| 156 | if data_info.source.radiation is None: |
|---|
| 157 | source_entry['radiation'] = _h5_string('neutron') |
|---|
| 158 | else: |
|---|
| 159 | source_entry['radiation'] = _h5_string(data_info.source.radiation) |
|---|
| 160 | if data_info.source.beam_shape is not None: |
|---|
| 161 | source_entry['beam_shape'] = _h5_string(data_info.source.beam_shape) |
|---|
| 162 | wavelength_keys = { 'wavelength': 'incident_wavelength', |
|---|
| 163 | 'wavelength_min':'wavelength_min', |
|---|
| 164 | 'wavelength_max': 'wavelength_max', |
|---|
| 165 | 'wavelength_spread': 'incident_wavelength_spread' } |
|---|
| 166 | for sasname, nxname in wavelength_keys.items(): |
|---|
| 167 | value = getattr(data_info.source, sasname) |
|---|
| 168 | units = getattr(data_info.source, sasname + '_unit') |
|---|
| 169 | if value is not None: |
|---|
| 170 | source_entry[nxname] = _h5_float(value) |
|---|
| 171 | source_entry[nxname].attrs['units'] = units |
|---|
| 172 | _write_h5_vector(source_entry, data_info.source.beam_size, |
|---|
| 173 | names=['beam_size_x', 'beam_size_y'], |
|---|
| 174 | units=data_info.source.beam_size_unit, write_fn=_write_h5_float) |
|---|
| 175 | |
|---|
| 176 | # Collimation metadata |
|---|
| 177 | if len(data_info.collimation) > 0: |
|---|
| 178 | for i, coll_info in enumerate(data_info.collimation): |
|---|
| 179 | collimation_entry = instrument_entry.create_group( |
|---|
| 180 | 'sascollimation{0:0=2d}'.format(i + 1)) |
|---|
| 181 | collimation_entry.attrs['canSAS_class'] = 'SAScollimation' |
|---|
| 182 | if coll_info.length is not None: |
|---|
| 183 | _write_h5_float(collimation_entry, coll_info.length, 'SDD') |
|---|
| 184 | collimation_entry['SDD'].attrs['units'] =\ |
|---|
| 185 | coll_info.length_unit |
|---|
| 186 | if coll_info.name is not None: |
|---|
| 187 | collimation_entry['name'] = _h5_string(coll_info.name) |
|---|
| 188 | else: |
|---|
| 189 | # Create a blank one - at least 1 collimation required by format |
|---|
| 190 | instrument_entry.create_group('sascollimation01') |
|---|
| 191 | |
|---|
| 192 | # Detector metadata |
|---|
| 193 | if len(data_info.detector) > 0: |
|---|
| 194 | i = 1 |
|---|
| 195 | for i, det_info in enumerate(data_info.detector): |
|---|
| 196 | detector_entry = instrument_entry.create_group( |
|---|
| 197 | 'sasdetector{0:0=2d}'.format(i + 1)) |
|---|
| 198 | detector_entry.attrs['canSAS_class'] = 'SASdetector' |
|---|
| 199 | if det_info.distance is not None: |
|---|
| 200 | _write_h5_float(detector_entry, det_info.distance, 'SDD') |
|---|
| 201 | detector_entry['SDD'].attrs['units'] =\ |
|---|
| 202 | det_info.distance_unit |
|---|
| 203 | if det_info.name is not None: |
|---|
| 204 | detector_entry['name'] = _h5_string(det_info.name) |
|---|
| 205 | else: |
|---|
| 206 | detector_entry['name'] = _h5_string('') |
|---|
| 207 | if det_info.slit_length is not None: |
|---|
| 208 | _write_h5_float(detector_entry, det_info.slit_length, |
|---|
| 209 | 'slit_length') |
|---|
| 210 | detector_entry['slit_length'].attrs['units'] =\ |
|---|
| 211 | det_info.slit_length_unit |
|---|
| 212 | _write_h5_vector(detector_entry, det_info.offset) |
|---|
| 213 | # NXcanSAS doesn't save information about pitch, only roll |
|---|
| 214 | # and yaw. The _write_h5_vector method writes vector.y, but we |
|---|
| 215 | # need to write vector.z for yaw |
|---|
| 216 | det_info.orientation.y = det_info.orientation.z |
|---|
| 217 | _write_h5_vector(detector_entry, det_info.orientation, |
|---|
| 218 | names=['polar_angle', 'azimuthal_angle']) |
|---|
| 219 | _write_h5_vector(detector_entry, det_info.beam_center, |
|---|
| 220 | names=['beam_center_x', 'beam_center_y'], |
|---|
| 221 | write_fn=_write_h5_float, units=det_info.beam_center_unit) |
|---|
| 222 | _write_h5_vector(detector_entry, det_info.pixel_size, |
|---|
| 223 | names=['x_pixel_size', 'y_pixel_size'], |
|---|
| 224 | write_fn=_write_h5_float, units=det_info.pixel_size_unit) |
|---|
| 225 | else: |
|---|
| 226 | # Create a blank one - at least 1 detector required by format |
|---|
| 227 | detector_entry = instrument_entry.create_group('sasdetector01') |
|---|
| 228 | detector_entry.attrs['canSAS_class'] = 'SASdetector' |
|---|
| 229 | detector_entry.attrs['name'] = '' |
|---|
| 230 | |
|---|
| 231 | # Process meta data |
|---|
| 232 | for i, process in enumerate(data_info.process): |
|---|
| 233 | process_entry = sasentry.create_group('sasprocess{0:0=2d}'.format( |
|---|
| 234 | i + 1)) |
|---|
| 235 | process_entry.attrs['canSAS_class'] = 'SASprocess' |
|---|
| 236 | if process.name: |
|---|
| 237 | name = _h5_string(process.name) |
|---|
| 238 | process_entry.create_dataset('name', data=name) |
|---|
| 239 | if process.date: |
|---|
| 240 | date = _h5_string(process.date) |
|---|
| 241 | process_entry.create_dataset('date', data=date) |
|---|
| 242 | if process.description: |
|---|
| 243 | desc = _h5_string(process.description) |
|---|
| 244 | process_entry.create_dataset('description', data=desc) |
|---|
| 245 | for j, term in enumerate(process.term): |
|---|
| 246 | # Don't save empty terms |
|---|
| 247 | if term: |
|---|
| 248 | h5_term = _h5_string(term) |
|---|
| 249 | process_entry.create_dataset('term{0:0=2d}'.format( |
|---|
| 250 | j + 1), data=h5_term) |
|---|
| 251 | for j, note in enumerate(process.notes): |
|---|
| 252 | # Don't save empty notes |
|---|
| 253 | if note: |
|---|
| 254 | h5_note = _h5_string(note) |
|---|
| 255 | process_entry.create_dataset('note{0:0=2d}'.format( |
|---|
| 256 | j + 1), data=h5_note) |
|---|
| 257 | |
|---|
| 258 | # Transmission Spectrum |
|---|
| 259 | for i, trans in enumerate(data_info.trans_spectrum): |
|---|
| 260 | trans_entry = sasentry.create_group( |
|---|
| 261 | 'sastransmission_spectrum{0:0=2d}'.format(i + 1)) |
|---|
| 262 | trans_entry.attrs['canSAS_class'] = 'SAStransmission_spectrum' |
|---|
| 263 | trans_entry.attrs['signal'] = 'T' |
|---|
| 264 | trans_entry.attrs['T_axes'] = 'T' |
|---|
| 265 | trans_entry.attrs['name'] = trans.name |
|---|
| 266 | if trans.timestamp is not '': |
|---|
| 267 | trans_entry.attrs['timestamp'] = trans.timestamp |
|---|
| 268 | transmission = trans_entry.create_dataset('T', |
|---|
| 269 | data=trans.transmission) |
|---|
| 270 | transmission.attrs['unertainties'] = 'Tdev' |
|---|
| 271 | trans_entry.create_dataset('Tdev', |
|---|
| 272 | data=trans.transmission_deviation) |
|---|
| 273 | trans_entry.create_dataset('lambda', data=trans.wavelength) |
|---|
| 274 | |
|---|
| 275 | note_entry = sasentry.create_group('sasnote'.format(i)) |
|---|
| 276 | note_entry.attrs['canSAS_class'] = 'SASnote' |
|---|
| 277 | notes = None |
|---|
| 278 | if len(data_info.notes) > 1: |
|---|
| 279 | notes = [np.string_(n) for n in data_info.notes] |
|---|
| 280 | notes = np.array(notes) |
|---|
| 281 | elif data_info.notes != []: |
|---|
| 282 | notes = _h5_string(data_info.notes[0]) |
|---|
| 283 | if notes is not None: |
|---|
| 284 | note_entry.create_dataset('SASnote', data=notes) |
|---|
| 285 | |
|---|
| 286 | f.close() |
|---|
| 287 | |
|---|
| 288 | def _write_1d_data(self, data_obj, data_entry): |
|---|
| 289 | """ |
|---|
| 290 | Writes the contents of a Data1D object to a SASdata h5py Group |
|---|
| 291 | |
|---|
| 292 | :param data_obj: A Data1D object to write to the file |
|---|
| 293 | :param data_entry: A h5py Group object representing the SASdata |
|---|
| 294 | """ |
|---|
| 295 | data_entry.attrs['signal'] = 'I' |
|---|
| 296 | data_entry.attrs['I_axes'] = 'Q' |
|---|
| 297 | data_entry.attrs['Q_indices'] = [0] |
|---|
| 298 | q_entry = data_entry.create_dataset('Q', data=data_obj.x) |
|---|
| 299 | q_entry.attrs['units'] = data_obj.x_unit |
|---|
| 300 | i_entry = data_entry.create_dataset('I', data=data_obj.y) |
|---|
| 301 | i_entry.attrs['units'] = data_obj.y_unit |
|---|
| 302 | if data_obj.dy is not None: |
|---|
| 303 | i_entry.attrs['uncertainties'] = 'Idev' |
|---|
| 304 | i_dev_entry = data_entry.create_dataset('Idev', data=data_obj.dy) |
|---|
| 305 | i_dev_entry.attrs['units'] = data_obj.y_unit |
|---|
| 306 | if data_obj.dx is not None: |
|---|
| 307 | q_entry.attrs['resolutions'] = 'dQ' |
|---|
| 308 | dq_entry = data_entry.create_dataset('dQ', data=data_obj.dx) |
|---|
| 309 | dq_entry.attrs['units'] = data_obj.x_unit |
|---|
| 310 | elif data_obj.dxl is not None: |
|---|
| 311 | q_entry.attrs['resolutions'] = ['dQl','dQw'] |
|---|
| 312 | dql_entry = data_entry.create_dataset('dQl', data=data_obj.dxl) |
|---|
| 313 | dql_entry.attrs['units'] = data_obj.x_unit |
|---|
| 314 | dqw_entry = data_entry.create_dataset('dQw', data=data_obj.dxw) |
|---|
| 315 | dqw_entry.attrs['units'] = data_obj.x_unit |
|---|
| 316 | |
|---|
| 317 | def _write_2d_data(self, data, data_entry): |
|---|
| 318 | """ |
|---|
| 319 | Writes the contents of a Data2D object to a SASdata h5py Group |
|---|
| 320 | |
|---|
| 321 | :param data: A Data2D object to write to the file |
|---|
| 322 | :param data_entry: A h5py Group object representing the SASdata |
|---|
| 323 | """ |
|---|
| 324 | data_entry.attrs['signal'] = 'I' |
|---|
| 325 | data_entry.attrs['I_axes'] = 'Qx,Qy' |
|---|
| 326 | data_entry.attrs['Q_indices'] = [0,1] |
|---|
| 327 | |
|---|
| 328 | (n_rows, n_cols) = (len(data.y_bins), len(data.x_bins)) |
|---|
| 329 | |
|---|
| 330 | if (n_rows == 0 and n_cols == 0) or (n_cols*n_rows != data.data.size): |
|---|
| 331 | # Calculate rows and columns, assuming detector is square |
|---|
| 332 | # Same logic as used in PlotPanel.py _get_bins |
|---|
| 333 | n_cols = int(np.floor(np.sqrt(len(data.qy_data)))) |
|---|
| 334 | n_rows = int(np.floor(len(data.qy_data) / n_cols)) |
|---|
| 335 | |
|---|
| 336 | if n_rows * n_cols != len(data.qy_data): |
|---|
| 337 | raise ValueError("Unable to calculate dimensions of 2D data") |
|---|
| 338 | |
|---|
| 339 | intensity = np.reshape(data.data, (n_rows, n_cols)) |
|---|
| 340 | qx = np.reshape(data.qx_data, (n_rows, n_cols)) |
|---|
| 341 | qy = np.reshape(data.qy_data, (n_rows, n_cols)) |
|---|
| 342 | |
|---|
| 343 | i_entry = data_entry.create_dataset('I', data=intensity) |
|---|
| 344 | i_entry.attrs['units'] = data.I_unit |
|---|
| 345 | qx_entry = data_entry.create_dataset('Qx', data=qx) |
|---|
| 346 | qx_entry.attrs['units'] = data.Q_unit |
|---|
| 347 | qy_entry = data_entry.create_dataset('Qy', data=qy) |
|---|
| 348 | qy_entry.attrs['units'] = data.Q_unit |
|---|
| 349 | if (data.err_data is not None |
|---|
| 350 | and not all(v is None for v in data.err_data)): |
|---|
| 351 | d_i = np.reshape(data.err_data, (n_rows, n_cols)) |
|---|
| 352 | i_entry.attrs['uncertainties'] = 'Idev' |
|---|
| 353 | i_dev_entry = data_entry.create_dataset('Idev', data=d_i) |
|---|
| 354 | i_dev_entry.attrs['units'] = data.I_unit |
|---|
| 355 | if (data.dqx_data is not None |
|---|
| 356 | and not all(v is None for v in data.dqx_data)): |
|---|
| 357 | qx_entry.attrs['resolutions'] = 'dQx' |
|---|
| 358 | dqx_entry = data_entry.create_dataset('dQx', data=data.dqx_data) |
|---|
| 359 | dqx_entry.attrs['units'] = data.Q_unit |
|---|
| 360 | if (data.dqy_data is not None |
|---|
| 361 | and not all(v is None for v in data.dqy_data)): |
|---|
| 362 | qy_entry.attrs['resolutions'] = 'dQy' |
|---|
| 363 | dqy_entry = data_entry.create_dataset('dQy', data=data.dqy_data) |
|---|
| 364 | dqy_entry.attrs['units'] = data.Q_unit |
|---|
| 365 | if data.mask is not None and not all(v is None for v in data.mask): |
|---|
| 366 | data_entry.attrs['mask'] = "mask" |
|---|
| 367 | mask = np.invert(np.asarray(data.mask, dtype=bool)) |
|---|
| 368 | data_entry.create_dataset('mask', data=mask) |
|---|
