ascii_reader.py @ d59cb8c

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-4.2.2release_4.0.1ticket-1009ticket-1094-headlessticket-1242-2d-resolutionticket-1243ticket-1249ticket885unittest-saveload

Last change on this file since d59cb8c was ef9d209, checked in by Jae Cho <jhjcho@…>, 13 years ago
added csv format: but read floats only but not strings
Property mode set to `100644`
File size: 16.0 KB

Line
1
2
3	############################################################################
4	#This software was developed by the University of Tennessee as part of the
5	#Distributed Data Analysis of Neutron Scattering Experiments (DANSE)
6	#project funded by the US National Science Foundation.
7	#If you use DANSE applications to do scientific research that leads to
8	#publication, we ask that you acknowledge the use of the software with the
9	#following sentence:
10	#This work benefited from DANSE software developed under NSF award DMR-0520547.
11	#copyright 2008, University of Tennessee
12	#############################################################################
13
14
15	import numpy
16	import os
17	from sans.dataloader.data_info import Data1D
18
19	# Check whether we have a converter available
20	has_converter = True
21	try:
22	from data_util.nxsunit import Converter
23	except:
24	has_converter = False
25	_ZERO = 1e-16
26
27	class Reader:
28	"""
29	Class to load ascii files (2, 3 or 4 columns).
30	"""
31	## File type
32	type_name = "ASCII"
33
34	## Wildcards
35	type = ["ASCII files (.txt)\|.txt",
36	"ASCII files (.dat)\|.dat",
37	"ASCII files (.abs)\|.abs",
38	"CSV files (.csv)\|.csv"]
39	## List of allowed extensions
40	ext = ['.txt', '.TXT', '.dat', '.DAT', '.abs', '.ABS', 'csv', 'CSV']
41
42	## Flag to bypass extension check
43	allow_all = True
44
45	def read(self, path):
46	"""
47	Load data file
48
49	:param path: file path
50
51	:return: Data1D object, or None
52
53	:raise RuntimeError: when the file can't be opened
54	:raise ValueError: when the length of the data vectors are inconsistent
55	"""
56	if os.path.isfile(path):
57	basename = os.path.basename(path)
58	_, extension = os.path.splitext(basename)
59	if self.allow_all or extension.lower() in self.ext:
60	try:
61	input_f = open(path,'r')
62	except :
63	raise RuntimeError, "ascii_reader: cannot open %s" % path
64	buff = input_f.read()
65	lines = buff.split('\n')
66
67	#Jae could not find python universal line spliter:
68	#keep the below for now
69	# some ascii data has \r line separator,
70	# try it when the data is on only one long line
71	if len(lines) < 2 :
72	lines = buff.split('\r')
73
74	x = numpy.zeros(0)
75	y = numpy.zeros(0)
76	dy = numpy.zeros(0)
77	dx = numpy.zeros(0)
78
79	#temp. space to sort data
80	tx = numpy.zeros(0)
81	ty = numpy.zeros(0)
82	tdy = numpy.zeros(0)
83	tdx = numpy.zeros(0)
84
85	output = Data1D(x, y, dy=dy, dx=dx)
86	self.filename = output.filename = basename
87
88	data_conv_q = None
89	data_conv_i = None
90
91	if has_converter == True and output.x_unit != '1/A':
92	data_conv_q = Converter('1/A')
93	# Test it
94	data_conv_q(1.0, output.x_unit)
95
96	if has_converter == True and output.y_unit != '1/cm':
97	data_conv_i = Converter('1/cm')
98	# Test it
99	data_conv_i(1.0, output.y_unit)
100
101
102	# The first good line of data will define whether
103	# we have 2-column or 3-column ascii
104	has_error_dx = None
105	has_error_dy = None
106
107	#Initialize counters for data lines and header lines.
108	is_data = False #Has more than 5 lines
109	# More than "5" lines of data is considered as actual
110	# data unless that is the only data
111	mum_data_lines = 5
112	# To count # of current data candidate lines
113	i = -1
114	# To count total # of previous data candidate lines
115	i1 = -1
116	# To count # of header lines
117	j = -1
118	# Helps to count # of header lines
119	j1 = -1
120	#minimum required number of columns of data; ( <= 4).
121	lentoks = 2
122	for line in lines:
123	toks = line.split(',')
124	if len(toks) < 2:
125	toks = line.split()
126	try:
127	#Make sure that all columns are numbers.
128	for colnum in range(len(toks)):
129	float(toks[colnum])
130
131	_x = float(toks[0])
132	_y = float(toks[1])
133
134	#Reset the header line counters
135	if j == j1:
136	j = 0
137	j1 = 0
138
139	if i > 1:
140	is_data = True
141
142	if data_conv_q is not None:
143	_x = data_conv_q(_x, units=output.x_unit)
144
145	if data_conv_i is not None:
146	_y = data_conv_i(_y, units=output.y_unit)
147
148	# If we have an extra token, check
149	# whether it can be interpreted as a
150	# third column.
151	_dy = None
152	if len(toks) > 2:
153	try:
154	_dy = float(toks[2])
155
156	if data_conv_i is not None:
157	_dy = data_conv_i(_dy, units=output.y_unit)
158
159	except:
160	# The third column is not a float, skip it.
161	pass
162
163	# If we haven't set the 3rd column
164	# flag, set it now.
165	if has_error_dy == None:
166	has_error_dy = False if _dy == None else True
167
168	#Check for dx
169	_dx = None
170	if len(toks) > 3:
171	try:
172	_dx = float(toks[3])
173
174	if data_conv_i is not None:
175	_dx = data_conv_i(_dx, units=output.x_unit)
176
177	except:
178	# The 4th column is not a float, skip it.
179	pass
180
181	# If we haven't set the 3rd column
182	# flag, set it now.
183	if has_error_dx == None:
184	has_error_dx = False if _dx == None else True
185
186	#After talked with PB, we decided to take care of only
187	# 4 columns of data for now.
188	#number of columns in the current line
189	#To remember the # of columns in the current
190	#line of data
191	new_lentoks = len(toks)
192
193	#If the previous columns not equal to the current,
194	#mark the previous as non-data and reset the dependents.
195	if lentoks != new_lentoks :
196	if is_data == True:
197	break
198	else:
199	i = -1
200	i1 = 0
201	j = -1
202	j1 = -1
203
204
205	#Delete the previously stored lines of data candidates
206	# if is not data.
207	if i < 0 and -1 < i1 < mum_data_lines and \
208	is_data == False:
209	try:
210	x = numpy.zeros(0)
211	y = numpy.zeros(0)
212	except:
213	pass
214
215	x = numpy.append(x, _x)
216	y = numpy.append(y, _y)
217
218	if has_error_dy == True:
219	#Delete the previously stored lines of
220	# data candidates if is not data.
221	if i < 0 and -1 < i1 < mum_data_lines and \
222	is_data == False:
223	try:
224	dy = numpy.zeros(0)
225	except:
226	pass
227	dy = numpy.append(dy, _dy)
228
229	if has_error_dx == True:
230	#Delete the previously stored lines of
231	# data candidates if is not data.
232	if i < 0 and -1 < i1 < mum_data_lines and \
233	is_data == False:
234	try:
235	dx = numpy.zeros(0)
236	except:
237	pass
238	dx = numpy.append(dx, _dx)
239
240	#Same for temp.
241	#Delete the previously stored lines of data candidates
242	# if is not data.
243	if i < 0 and -1 < i1 < mum_data_lines and\
244	is_data == False:
245	try:
246	tx = numpy.zeros(0)
247	ty = numpy.zeros(0)
248	except:
249	pass
250
251	tx = numpy.append(tx, _x)
252	ty = numpy.append(ty, _y)
253
254	if has_error_dy == True:
255	#Delete the previously stored lines of
256	# data candidates if is not data.
257	if i < 0 and -1 < i1 < mum_data_lines and \
258	is_data == False:
259	try:
260	tdy = numpy.zeros(0)
261	except:
262	pass
263	tdy = numpy.append(tdy, _dy)
264	if has_error_dx == True:
265	#Delete the previously stored lines of
266	# data candidates if is not data.
267	if i < 0 and -1 < i1 < mum_data_lines and \
268	is_data == False:
269	try:
270	tdx = numpy.zeros(0)
271	except:
272	pass
273	tdx = numpy.append(tdx, _dx)
274
275	#reset i1 and flag lentoks for the next
276	if lentoks < new_lentoks:
277	if is_data == False:
278	i1 = -1
279	#To remember the # of columns on the current line
280	# for the next line of data
281	lentoks = len(toks)
282
283	#Reset # of header lines and counts #
284	# of data candidate lines
285	if j == 0 and j1 == 0:
286	i1 = i + 1
287	i += 1
288	except:
289
290	# It is data and meet non - number, then stop reading
291	if is_data == True:
292	break
293	lentoks = 2
294	#Counting # of header lines
295	j += 1
296	if j == j1 + 1:
297	j1 = j
298	else:
299	j = -1
300	#Reset # of lines of data candidates
301	i = -1
302
303	# Couldn't parse this line, skip it
304	pass
305
306	input_f.close()
307	# Sanity check
308	if has_error_dy == True and not len(y) == len(dy):
309	msg = "ascii_reader: y and dy have different length"
310	raise RuntimeError, msg
311	if has_error_dx == True and not len(x) == len(dx):
312	msg = "ascii_reader: y and dy have different length"
313	raise RuntimeError, msg
314	# If the data length is zero, consider this as
315	# though we were not able to read the file.
316	if len(x) == 0:
317	raise RuntimeError, "ascii_reader: could not load file"
318
319	#Let's re-order the data to make cal.
320	# curve look better some cases
321	ind = numpy.lexsort((ty, tx))
322	for i in ind:
323	x[i] = tx[ind[i]]
324	y[i] = ty[ind[i]]
325	if has_error_dy == True:
326	dy[i] = tdy[ind[i]]
327	if has_error_dx == True:
328	dx[i] = tdx[ind[i]]
329	# Zeros in dx, dy
330	if has_error_dx:
331	dx[dx==0] = _ZERO
332	if has_error_dy:
333	dy[dy==0] = _ZERO
334	#Data
335	output.x = x[x!=0]
336	output.y = y[x!=0]
337	output.dy = dy[x!=0] if has_error_dy == True else numpy.zeros(len(output.y))
338	output.dx = dx[x!=0] if has_error_dx == True else numpy.zeros(len(output.x))
339
340	if data_conv_q is not None:
341	output.xaxis("\\rm{Q}", output.x_unit)
342	else:
343	output.xaxis("\\rm{Q}", 'A^{-1}')
344	if data_conv_i is not None:
345	output.yaxis("\\rm{Intensity}", output.y_unit)
346	else:
347	output.yaxis("\\rm{Intensity}","cm^{-1}")
348
349	# Store loading process information
350	output.meta_data['loader'] = self.type_name
351
352	return output
353
354	else:
355	raise RuntimeError, "%s is not a file" % path
356	return None
357
358	if __name__ == "__main__":
359	reader = Reader()
360	#print reader.read("../test/test_3_columns.txt")
361	print reader.read("../test/empty.txt")
362
363
364

Note: See TracBrowser for help on using the repository browser.

SasView

source: sasview/sansdataloader/src/sans/dataloader/readers/ascii_reader.py @ d59cb8c

Download in other formats: