Newer
Older
'''
A Variable is a full ICARTT variable with name, units, scale and missing value.
'''
'''
Return variable description string as it appears in an ICARTT file
'''
return self.splitChar.join( [self.name, self.units, self.units ] )
def __init__(self, name, units, scale=1.0, miss=-9999999):
'''
An ICARTT dataset that can be created from scratch or read from a file,
manipulated, and then written to a file.
'''
if self.format == 2110:
total += self.nauxvar + 5
return total
@property
def ndvar(self):
return len(self.DVAR)
@property
def nauxvar(self):
return len(self.AUXVAR)
@property
def nvar(self):
return self.ndvar + 1
@property
def nscom(self):
return len(self.SCOM)
@property
def nncom(self):
return len(self.NCOM)
@property
def VAR(self):
return [ self.IVAR ] + self.DVAR
@property
def varnames(self):
return [ x.name for x in self.VAR ]
def units(self, name):
res = [ x.units for x in self.VAR if x.name == name ]
if len(res) is 0:
res = [ '' ]
return res[0]
def index(self, name):
res = [ i for i, x in enumerate(self.VAR) if x.name == name ]
if len(res) is 0:
res = [ -1 ]
return res[0]
def write(self, f=sys.stdout):
def prnt(txt):
f.write(str(txt) + "\n")
# Number of lines in header, file format index (most files use 1001) - comma delimited.
prnt("{:d}, {:d}".format(self.nheader, self.format))
# PI last name, first name/initial.
prnt(self.PI)
# Organization/affiliation of PI.
prnt(self.organization)
# Data source description (e.g., instrument name, platform name, model name, etc.).
prnt(self.dataSource)
# Mission name (usually the mission acronym).
prnt(self.mission)
# File volume number, number of file volumes (these integer values are used when the data require more than one file per day; for data that require only one file these values are set to 1, 1) - comma delimited.
# UTC date when data begin, UTC date of data reduction or revision - comma delimited (yyyy, mm, dd, yyyy, mm, dd).
prnt(self.splitChar.join([ datetime.datetime.strftime(x, "%Y, %m, %d") for x in [ self.dateValid, self.dateRevised ] ]))
# Data Interval (This value describes the time spacing (in seconds) between consecutive data records. It is the (constant) interval between values of the independent variable. For 1 Hz data the data interval value is 1 and for 10 Hz data the value is 0.1. All intervals longer than 1 second must be reported as Start and Stop times, and the Data Interval value is set to 0. The Mid-point time is required when it is not at the average of Start and Stop times. For additional information see Section 2.5 below.).
prnt("0")
# Description or name of independent variable (This is the name chosen for the start time. It always refers to the number of seconds UTC from the start of the day on which measurements began. It should be noted here that the independent variable should monotonically increase even when crossing over to a second day.).
prnt(self.IVAR.desc)
if self.format == 2110:
# Description or name of independent (bound) variable (This is the name chosen for the start time. It always refers to the number of seconds UTC from the start of the day on which measurements began. It should be noted here that the independent variable should monotonically increase even when crossing over to a second day.).
prnt(self.IBVAR.desc)
# Number of variables (Integer value showing the number of dependent variables: the total number of columns of data is this value plus one.).
prnt(self.ndvar)
# Scale factors (1 for most cases, except where grossly inconvenient) - comma delimited.
prnt(self.splitChar.join( [ "{:6.3f}".format(x.scale) for x in self.DVAR ]))
# Missing data indicators (This is -9999 (or -99999, etc.) for any missing data condition, except for the main time (independent) variable which is never missing) - comma delimited.
prnt(self.splitChar.join( [ "{:d}".format(x.miss) for x in self.DVAR ]))
# Variable names and units (Short variable name and units are required, and optional long descriptive name, in that order, and separated by commas. If the variable is unitless, enter the keyword "none" for its units. Each short variable name and units (and optional long name) are entered on one line. The short variable name must correspond exactly to the name used for that variable as a column header, i.e., the last header line prior to start of data.).
nul = [ prnt(x.desc) for x in self.DVAR ]
if self.format == 2110:
# Number of variables (Integer value showing the number of dependent variables: the total number of columns of data is this value plus one.).
prnt(self.nauxvar)
# Scale factors (1 for most cases, except where grossly inconvenient) - comma delimited.
prnt(self.splitChar.join( [ "{:6.3f}".format(x.scale) for x in self.AUXVAR ]))
# Missing data indicators (This is -9999 (or -99999, etc.) for any missing data condition, except for the main time (independent) variable which is never missing) - comma delimited.
prnt(self.splitChar.join( [ "{:d}".format(x.miss) for x in self.AUXVAR ]))
# Variable names and units (Short variable name and units are required, and optional long descriptive name, in that order, and separated by commas. If the variable is unitless, enter the keyword "none" for its units. Each short variable name and units (and optional long name) are entered on one line. The short variable name must correspond exactly to the name used for that variable as a column header, i.e., the last header line prior to start of data.).
nul = [ prnt(x.desc) for x in self.AUXVAR ]
# Number of SPECIAL comment lines (Integer value indicating the number of lines of special comments, NOT including this line.).
prnt("{:d}".format(self.nscom))
# Special comments (Notes of problems or special circumstances unique to this file. An example would be comments/problems associated with a particular flight.).
nul = [ prnt(x) for x in self.SCOM ]
# Number of Normal comments (i.e., number of additional lines of SUPPORTING information: Integer value indicating the number of lines of additional information, NOT including this line.).
prnt("{:d}".format(self.nncom))
# Normal comments (SUPPORTING information: This is the place for investigators to more completely describe the data and measurement parameters. The supporting information structure is described below as a list of key word: value pairs. Specifically include here information on the platform used, the geo-location of data, measurement technique, and data revision comments. Note the non-optional information regarding uncertainty, the upper limit of detection (ULOD) and the lower limit of detection (LLOD) for each measured variable. The ULOD and LLOD are the values, in the same units as the measurements that correspond to the flags -7777s and -8888s within the data, respectively. The last line of this section should contain all the short variable names on one line. The key words in this section are written in BOLD below and must appear in this section of the header along with the relevant data listed after the colon. For key words where information is not needed or applicable, simply enter N/A.).
nul = [ prnt(x) for x in self.NCOM ]
nul = [ prnt(self.splitChar.join( [ str(y) for y in x ] )) for x in self.data ]
'''
Create ICARTT-compliant file name based on the information contained in the dataset
'''
return self.dataID + "_" +self.locationID + "_" +datetime.datetime.strftime(self.dateValid, '%Y%m%d') + "_" +"R" + self.revision + ".ict"
# sanitize function
def __readline(self, do_split=True):
dmp = self.input_fhandle.readline().replace('\n', '').replace('\r','')
if do_split:
dmp = [word.strip(' ') for word in dmp.split(self.splitChar) ]
'''
Read the ICARTT header (from file)
'''
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
if self.input_fhandle.closed:
self.input_fhandle = open(self.input_fhandle.name)
# line 1 - Number of lines in header, file format index (most files use
# 1001) - comma delimited.
self.format = int(self.__readline()[1])
# line 2 - PI last name, first name/initial.
self.PI = self.__readline(do_split=False)
# line 3 - Organization/affiliation of PI.
self.organization = self.__readline(do_split=False)
# line 4 - Data source description (e.g., instrument name, platform name,
# model name, etc.).
self.dataSource = self.__readline(do_split=False)
# line 5 - Mission name (usually the mission acronym).
self.mission = self.__readline(do_split=False)
# line 6 - File volume number, number of file volumes (these integer values
# are used when the data require more than one file per day; for data that
# require only one file these values are set to 1, 1) - comma delimited.
dmp = self.__readline()
self.VOL = int(dmp[0])
self.NVOL = int(dmp[1])
# line 7 - UTC date when data begin, UTC date of data reduction or revision
# - comma delimited (yyyy, mm, dd, yyyy, mm, dd).
dmp = self.__readline()
self.dateValid = datetime.datetime.strptime("".join([ "{:s}".format(x) for x in dmp[0:3] ]), '%Y%m%d')
self.dateRevised = datetime.datetime.strptime("".join([ "{:s}".format(x) for x in dmp[3:6] ]), '%Y%m%d')
# line 8 - Data Interval (This value describes the time spacing (in seconds)
# between consecutive data records. It is the (constant) interval between
# values of the independent variable. For 1 Hz data the data interval value
# is 1 and for 10 Hz data the value is 0.1. All intervals longer than 1
# second must be reported as Start and Stop times, and the Data Interval
# value is set to 0. The Mid-point time is required when it is not at the
# average of Start and Stop times. For additional information see Section
# 2.5 below.).
self.dataInterval = int(self.__readline()[0])
# line 9 - Description or name of independent variable (This is the name
# chosen for the start time. It always refers to the number of seconds UTC
# from the start of the day on which measurements began. It should be noted
# here that the independent variable should monotonically increase even when
# crossing over to a second day.
dmp = self.__readline()
self.IVAR = Variable(dmp[0], dmp[1])
# line 10 - Number of variables (Integer value showing the number of
# dependent variables: the total number of columns of data is this value
# plus one.).
ndvar = int(self.__readline()[0])
# line 11- Scale factors (1 for most cases, except where grossly
# inconvenient) - comma delimited.
dvscale = [ float(x) for x in self.__readline() ]
# line 12 - Missing data indicators (This is -9999 (or -99999, etc.) for
# any missing data condition, except for the main time (independent)
# variable which is never missing) - comma delimited.
dvmiss = [ x for x in self.__readline() ]
# no float casting here, as we need to do string comparison lateron when reading data...
# line 13 - Variable names and units (Short variable name and units are
# required, and optional long descriptive name, in that order, and separated
# by commas. If the variable is unitless, enter the keyword "none" for its
# units. Each short variable name and units (and optional long name) are
# entered on one line. The short variable name must correspond exactly to
# the name used for that variable as a column header, i.e., the last header
# line prior to start of data.).
dmp = self.__readline()
dvname = [ dmp[0] ]
dvunits = [ dmp[1] ]
for i in range(1, ndvar):
dmp = self.__readline()
dvname += [ dmp[0] ]
dvunits += [ dmp[1] ]
self.DVAR = [ Variable(name, unit, scale, miss) for name, unit, scale, miss in zip(dvname, dvunits, dvscale, dvmiss) ]
# line 14 + nvar - Number of SPECIAL comment lines (Integer value
# indicating the number of lines of special comments, NOT including this
# line.).
nscom = int(self.__readline()[0])
# line 15 + nvar - Special comments (Notes of problems or special
# circumstances unique to this file. An example would be comments/problems
# associated with a particular flight.).
self.SCOM = [ self.__readline(do_split=False) for i in range(0, nscom) ]
# line 16 + nvar + nscom - Number of Normal comments (i.e., number of
# additional lines of SUPPORTING information: Integer value indicating the
# number of lines of additional information, NOT including this line.).
nncom = int(self.__readline()[0])
# line 17 + nvar + nscom - Normal comments (SUPPORTING information: This is
# the place for investigators to more completely describe the data and
# measurement parameters. The supporting information structure is described
# below as a list of key word: value pairs. Specifically include here
# information on the platform used, the geo-location of data, measurement
# technique, and data revision comments. Note the non-optional information
# regarding uncertainty, the upper limit of detection (ULOD) and the lower
# limit of detection (LLOD) for each measured variable. The ULOD and LLOD
# are the values, in the same units as the measurements that correspond to
# the flags -7777's and -8888's within the data, respectively. The last line
# of this section should contain all the "short" variable names on one line.
# The key words in this section are written in BOLD below and must appear in
# this section of the header along with the relevant data listed after the
# colon. For key words where information is not needed or applicable, simply
# enter N/A.).
self.NCOM = [ self.__readline(do_split=False) for i in range(0, nncom) ]
self.input_fhandle.close()
def __nan_miss_float(self, raw):
return [ float(x.replace(self.VAR[i].miss, 'NaN')) for i, x in enumerate(raw) ]
def read_data(self):
'''
Read ICARTT data (from file)
'''
if self.input_fhandle.closed:
self.input_fhandle = open(self.input_fhandle.name)
nul = [ self.input_fhandle.readline() for i in xrange(self.nheader) ]
self.data = [ self.__nan_miss_float(line.split(self.splitChar)) for line in self.input_fhandle ]
self.input_fhandle.close()
def read_first_and_last(self):
'''
Read first and last ICARTT data line (from file). Useful for quick estimates e.g. of the time extent
of big ICARTT files, without having to read the whole thing, which would be slow.
'''
if self.input_fhandle.closed:
self.input_fhandle = open(self.input_fhandle.name)
nul = [ self.input_fhandle.readline() for i in xrange(self.nheader) ]
first = self.input_fhandle.readline()
self.data = [ self.__nan_miss_float(first.split(self.splitChar)) ]
for line in self.input_fhandle:
pass
last = line
self.data += [ self.__nan_miss_float(last.split(',')) ]
self.input_fhandle.close()
def read(self):
self.read_header()
self.read_data()
def __init__(self, f=None, loadData=False):
self.revision = '0'
self.dataID = 'dataID'
self.locationID = 'locationID'
self.PI = 'Mustermann, Martin'
self.organization = 'Musterinstitut'
self.dataSource = 'Musterdatenprodukt'
self.mission = 'MUSTEREX'
self.dateValid = datetime.datetime.today()
self.dateRevised = datetime.datetime.today()
self.dataInterval = 0
self.IVAR = Variable('Time_Start',
'seconds_from_0_hours_on_valid_date',
1.0, -9999999)
self.DVAR = [
Variable('Time_Stop',
'seconds_from_0_hours_on_valid_date',
self.splitChar = ','
# read data if f is not None
if f is not None:
if isinstance(f, (str, unicode)):
self.input_fhandle = open(f, 'r')
else:
self.input_fhandle = f
self.read_header()
if loadData:
self.read_data()