-
Notifications
You must be signed in to change notification settings - Fork 14
/
lsdb1.py
408 lines (331 loc) · 15.9 KB
/
lsdb1.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
from datetime import datetime
import time
import db_lib
from collections import Counter
import dateutil.parser
import logging
logger = logging.getLogger(__name__)
#12/19 skinner didn't write this. Maybe Annie. Ask Matt.
def validate_date(d):
"""
This will convert about any date to be isoformat
this will take '2017-jul-12', '12-jul-2017', 'Jul-12-2017' and convert to '2017-07-12T00:00:00'
'2017-07-12' or '2017-07-12T00' and convert to '2017-07-12T00:00:00'
**it will not know which is the day,mont in the 07-12-2017**
"""
try:
datetime.strptime(d, '%Y-%m-%dT%H:%M:%S')
except ValueError:
new_d = dateutil.parser.parse(d)
d = new_d.isoformat()
logger.info('new formated date {}'.format(d))
return d
def getValidTimeInterval(start_thuman, end_thuman = None):
"""
In order to limit our querys, get headers results for an interval of time.
:params start_thuman, required
:params end_thuman opt. default to now
format is friendlier than timestamp 'yyyy-mm-ddThh:mm' (hours 0-24)
:return valid start and end dates
"""
# validate and convert to isoformat
start_thuman = validate_date(start_thuman)
start_t = time.mktime(datetime.strptime(start_thuman, '%Y-%m-%dT%H:%M:%S').timetuple())
if end_thuman is None:
now = datetime.now()
end_t =time.mktime(now.timetuple())
else:
end_thuman = validate_date(end_thuman)
end_t = time.mktime(datetime.strptime(end_thuman, '%Y-%m-%dT%H:%M:%S').timetuple())
return start_t, end_t
def getResultsByTimeInterval(start_thuman, end_thuman = None):
"""
Get headers for results for an interval of time.
:params start_thuman, required
:params end_thuman opt. default to now
format is friendlier than timestamp 'yyyy-mm-ddThh:mm' (hours 0-24)
:return list of full headers: format list of dict
"""
start_t, end_t = getValidTimeInterval(start_thuman, end_thuman)
headers = list(db_lib.analysis_ref.find_analysis_header(time={'$lt': end_t, '$gte': start_t}))
return headers
def getColRequestsByTimeInterval(start_thuman, end_thuman = None):
"""
Get collection requests for an interval of time.
:params start_thuman, required
:params end_thuman opt. default to now
format is friendlier than timestamp 'yyyy-mm-ddThh:mm' (hours 0-24)
:return list of full headers: format list of dict
"""
start_t, end_t = getValidTimeInterval(start_thuman, end_thuman)
headers = list(db_lib.request_ref.find(time={'$lt': end_t, '$gte': start_t},request_type={'$in':['standard','vector']},priority=-1))
return headers
def getAllRequestsByTimeInterval(start_thuman, end_thuman = None):
"""
Get all requests for an interval of time.
:param start_thuman, required
:param end_thuman opt. default to now
format is friendlier than timestamp 'yyyy-mm-ddThh:mm' (hours 0-24)
:return list of full headers: format list of dict
"""
start_t, end_t = getValidTimeInterval(start_thuman, end_thuman)
headers = list(db_lib.request_ref.find(time={'$lt': end_t, '$gte': start_t},priority=-1))
return headers
def getMountCount(start_thuman, end_thuman, beamline):
reqs = getResultsByTimeInterval(start_thuman, end_thuman)
mountCount=0
currentSampID = ""
for i in range (0,len(reqs)):
try:
if (reqs[i]["beamline"] == beamline):
sampID = reqs[i]["sample"]
if (sampID != currentSampID):
mountCount+=1
currentSampID = sampID
except KeyError:
logger.error('sample field not found: %s' % reqs[i])
return mountCount
def printColRequestsByTimeInterval(start_thuman, end_thuman = None, fname = "sweeps.txt"):
reqs = getColRequestsByTimeInterval(start_thuman, end_thuman)
outfile = open(fname,"w+")
outfile.write("proposalID\tdirectory\tprefix\tprotocol\tnumFrames\texptime\timgWidth\tomegaStart\ttranslation\txvec\tyvec\tzvec\tsysTime\ttimestamp\n")
for i in range (0,len(reqs)):
propID = reqs[i]["proposalID"]
reqObj = reqs[i]['request_obj']
exptime = reqObj['exposure_time']
imgWidth = reqObj['img_width']
sweep_start = reqObj['sweep_start']
sweep_end = reqObj['sweep_end']
prefix = reqObj['file_prefix']
directory = reqObj['directory']
protocol = reqObj['protocol']
if (protocol == "vector"):
translation = reqObj['vectorParams']['trans_total']
xvec = reqObj['vectorParams']['x_vec']
yvec = reqObj['vectorParams']['y_vec']
zvec = reqObj['vectorParams']['z_vec']
else:
translation = 0.0
xvec = 0.0
yvec = 0.0
zvec = 0.0
numimages = (sweep_end-sweep_start)/imgWidth
systemTime = reqs[i]["time"]
time_s = time.ctime(systemTime)
line = str(propID) + "\t" + directory + "\t" + prefix + "\t" + protocol + "\t" + str(numimages) + "\t" + str(exptime) + "\t" + str(imgWidth) + "\t" + str(sweep_start) + "\t" + str(translation) + "\t" + str(xvec) + "\t" + str(yvec) + "\t" + str(zvec) + "\t" + str(systemTime) + "\t" + time_s + "\n"
outfile.write(line)
logger.info(line)
logger.info(str(len(reqs)) + " requests")
outfile.close()
def getRasterRequestsByTimeInterval(start_thuman, end_thuman = None):
"""
Get headers for raster requests for an interval of time.
:params start_thuman, required
:params end_thuman opt. default to now
format is friendlier than timestamp 'yyyy-mm-ddThh:mm' (hours 0-24)
:return list of full headers: format list of dict
"""
start_t, end_t = getValidTimeInterval(start_thuman, end_thuman)
headers = list(db_lib.request_ref.find(time={'$lt': end_t, '$gte': start_t},request_type='raster'))
return headers
def getAllRequestsByTimeInterval(start_thuman, end_thuman = None):
"""
Get headers for all requests for an interval of time.
:params start_thuman, required
:params end_thuman opt. default to now
format is friendlier than timestamp 'yyyy-mm-ddThh:mm' (hours 0-24)
:return list of full headers: format list of dict
"""
start_t, end_t = getValidTimeInterval(start_thuman, end_thuman)
headers = list(db_lib.request_ref.find(time={'$lt': end_t, '$gte': start_t},priority=-1))
return headers
def getSampleReqRes(reqid):
"""
this function is used in the getVisitSummary.
the requestid is present in the request and in the results
if you don't have a list of requestid , you can try :
reqid = db_lib.getRequestsBySampleID(value, active_only=False
:params reqid is a list of requestid per sample
"""
logger.info('**New sample**')
if reqid != []:
logger.info('{:12} {:15} {:10}'.format('ProposalID', 'Sample Name', 'Sample ID'))
for r in reqid:
if r is not None:
logger.info('{:_>50}'.format(''))
sample_name = db_lib.getSampleNamebyID(r['sample'])
logger.info('{:12} {:15} {:10}'.format(r['proposalID'], sample_name, r['uid']))
reqRes = db_lib.getRequestByID(str(r['uid']), active_only = False)
FASTDP = 'no' # default
try:
FASTDP == reqRes['request_obj']['fastDP']
except KeyError as e:
logger.error(" no FASTDP key " + repr(e))
logger.info('{:>20} {:6} {:12} {:12}{:6} {:>6}-{:6}'.format('requests', 'FastDP', 'wavelength A',
'DetDist mm','width (Deg)', 'Start','End' ))
logger.info('{:>20} {:<6} {:10} {:10}{:6} {:>10}-{:6}'.format(r['request_type'], FASTDP,
reqRes['request_obj']['wavelength'],
reqRes['request_obj']['detDist'],
reqRes['request_obj']['img_width'],
reqRes['request_obj']['sweep_start'],
reqRes['request_obj']['sweep_end']))
logger.info('{:*>20}'.format('Results'))
try:
reqres = db_lib.getResultsforRequest(r['uid'])
for rr in reqres:
logger.info('{:>20}'.format(rr['result_type']))
# eventually it might be important to show what kind of results was obtained from the raster
if rr['result_type'] == 'rasterResult':
logger.info('{:>38}'.format(rr['result_obj']['rasterCellResults']['type']))
except Exception as e:
logger.error('in 2nd exception')
logger.error('{:20}'.format('none'))
def getSamplesByTimeInterval(start_thuman, end_thuman = None):
"""
get a unique list of sampleids associated with a proposal 'collected on" during an interval of time i.e
no discremination about the result_type
will parse the headers from et ResultsByTimeInterval
return: a dict {'proposalid':[sampleid]}
"""
headers = getResultsByTimeInterval(start_thuman, end_thuman)
dict_ps = {}
for p in headers:
if p['proposalID']:
key = p['proposalID']
if p['proposalID'] is not None:
if p['sample'] is not None:
sample = p['sample']
dict_ps.setdefault(key, list(set([]))).append(sample)
num_samples = ''
for k, v in dict_ps.items():
# make unique list of sampleid per proposal
dict_ps[k] = list(set(v))
# get some stats
num_samples = len(list(set(v)))
logger.info('proposalid {} had {} samples'.format(k, num_samples))
num_proposal = len(dict_ps.keys())
proposals = dict_ps.keys()
data = {"num_proposal": num_proposal, "proposals": proposals, "dict_samples": dict_ps}
return dict_ps
def getVisitSummary(start_thuman, end_thuman = None):
"""
Getting a summary of request and Results on samples collected during a visit WIP
:param start_thuman: required start time for the visit , format 'yyyy-mm-ddThh:mm'
:param end_thuman: opt if None defaults to now
:return: for now a screen print out, format to be determined , file that we can send?
"""
sampleid_dict = getSamplesByTimeInterval(start_thuman, end_thuman) #this probably eliminates test samples.
for proposalid, sid in sampleid_dict.items():
logger.info("proposal : %s" % proposalid)
for value in sid:
# getting a list of all the requests for a sample
# todo need to limit for the period
reqid = db_lib.getRequestsBySampleID(value, active_only=False)
try:
sampleInfo = getSampleReqRes(reqid)
except Exception as e:
logger.error('problem with sample info ' + repr(e))
def getShortVisitSummary(start_thuman, end_thuman=None):
"""
Get more generic statistics, on the mx beamlines, get a printed summary of what was done
:param start_thuman:
:param end_thuman:
:return: a list of dict for each proposalid worked on. This is for both beamlines (amx fmx)
"""
# get the headers
sampleid_dict = getSamplesByTimeInterval(start_thuman, end_thuman)
visits = [] # will have the stats for a visit according to proposalid.
# headers for the print
logger.info('{:25} {:10} {:10} {:^10}{:^10} {:^10} {:^10} {:10}'.format('Name', 'proposalid', 'standard', 'raster', 'snapshot', 'vector', 'escan', 'sample ID'))
for pid, sid in sampleid_dict.items():
num_samples = (len(sid))
# define timestamps for now start-end
p_stime = time.time()
p_etime = 1451624400.0 # '2016-01-01T00:00'
num_datacol = 0
for value in sid:
# getting the name of the samples
sample_name = db_lib.getSampleNamebyID(value)
# getting a flavor of the request type
list_type =[]
srt = db_lib.getRequestsBySampleID(value, active_only=False) # at this point it's not discriminating with time
num_datacol += len(srt)
# find the request_type for each header
for i in srt:
try:
req_type = i['request_type']
except Exception as e:
req_type = 'none'
# set the time period for a proposal
start_thuman = validate_date(start_thuman)
start_t = time.mktime(datetime.strptime(start_thuman, '%Y-%m-%dT%H:%M:%S' ).timetuple())
if end_thuman is None:
now = datetime.now()
end_t =time.mktime(now.timetuple())
else:
end_thuman = validate_date(end_thuman)
end_t = time.mktime(datetime.strptime(end_thuman, '%Y-%m-%dT%H:%M:%S').timetuple())
if i['time'] >= start_t and i['time'] <= end_t:
# set the time period for a proposal
if i['time'] <= p_stime:
p_stime = i['time']
if i['time'] >= p_etime:
p_etime = i['time']
list_type.append(req_type)
p_stime2 = datetime.fromtimestamp(p_stime).strftime('%Y-%m-%dT%H:%M:%S')
p_etime2 = datetime.fromtimestamp(p_etime).strftime('%Y-%m-%dT%H:%M:%S')
visit = {'pid': pid, 'num_datacol': num_datacol, 'beamline': 'amx', 'num_samples': num_samples,'start': p_stime2, 'end': p_etime2}
list_type = Counter(list_type)
standard = ''
raster = ''
escan = ''
vector = ''
snapshot = ''
if 'standard' in list_type:
standard = list_type['standard']
if 'raster' in list_type:
raster = list_type['raster']
if 'vector' in list_type:
vector = list_type['vector']
if 'snapshot' in list_type:
snapshot = list_type['snapshot']
if 'escan' in list_type:
escan = list_type['escan']
logger.info('{:25} {:^10} {:^10} {:^10} {:^10} {:^10} {:^10}{:10}'.format(sample_name, pid, standard, raster, snapshot, vector, escan, value))
visits.append(visit)
logger.info('----------{} samples, {} requests from {} to {} '.format(num_samples, num_datacol, p_stime2, p_etime2 ))
return visits
def getRequestsbyProposalID(proposalID, active_only=False):
"""
:param proposalID required int, 6 digits
:return a list of dict (headers)
"""
params = {'proposalID': proposalID}
if active_only:
params['state'] = "active"
reqs = list(db_lib.request_ref.find(**params))
return reqs
def getSamplesbyProposalID(proposalID):
"""
since proposalids are not included with the samples, they need to have a request attached to them
so, query for request with that proposalid and then extract the sampleid
do we need to attach the owner to the sample?
:params proposalID in theorie from PASS str(6 digits)
:return list of sampleid
"""
reqs = getRequestsbyProposalID(proposalID, active_only=False)
if reqs !=[]:
list_sampleid = []
for i in reqs:
list_sampleid.append(i['sample'])
logger.info('proposal {} has {} samples '.format(proposalID, len(list_sampleid)))
data = { 'proposalID': proposalID, 'num_samples': len(list_sampleid), 'list_sampleid': list_sampleid}
return list_sampleid
# def getBeamlineforProposalID(proposalID):
# """
# """
# reqs = getRequestsbyProposalID(proposalID, active_only=False)
# if reqs !=[]:
# list_beamline = []
# for i in reqs:
# list_beamline.append(i['beamline'])