def __init__(self, headername, liststations, filename, outputdir):
self.filename = filename
self.outputdir = outputdir
aa = read_ukmo(headername, liststations, filename)
# unique IDs
ids = npunique(aa.csvdata['ID'])
for identifier in ids:
try:
lon = aa.stationdata[identifier.strip()]['longitude']
lat = aa.stationdata[identifier.strip()]['latitude']
elevation = aa.stationdata[identifier.strip()]['elevation']
except KeyError:
continue
if (lon == -99999) or (lat == -99999) or (elevation == -99999):
continue
# list of indices
idx = npwhere(aa.csvdata['ID'] == identifier)[0]
# extract all keys for selected station identifier
dataout = collections.defaultdict(list)
dataout = dict(
(k, nparray(aa.csvdata[k])[idx]) for k in aa.csvdata.keys())
stationid = dataout['ID'][0]
# remove variables from dictionary
dataout.pop('longitude', None)
dataout.pop('latitude', None)
dataout.pop('elevation', None)
dataout.pop('ID', None)
# create netcdf file
filename = self.define_output_file(stationid)
self.write_netcdf(filename, dataout, lon, lat, elevation)
def stellingwerf_pdm_theta(times, mags, errs, frequency,
binsize=0.05, minbin=9):
'''
This calculates the Stellingwerf PDM theta value at a test frequency.
'''
period = 1.0/frequency
fold_time = times[0]
phased = phase_magseries(times,
mags,
period,
fold_time,
wrap=False,
sort=True)
phases = phased['phase']
pmags = phased['mags']
bins = np.arange(0.0, 1.0, binsize)
nbins = bins.size
binnedphaseinds = npdigitize(phases, bins)
binvariances = []
binndets = []
goodbins = 0
for x in npunique(binnedphaseinds):
thisbin_inds = binnedphaseinds == x
thisbin_phases = phases[thisbin_inds]
thisbin_mags = pmags[thisbin_inds]
if thisbin_mags.size > minbin:
thisbin_variance = npvar(thisbin_mags,ddof=1)
binvariances.append(thisbin_variance)
binndets.append(thisbin_mags.size)
goodbins = goodbins + 1
# now calculate theta
binvariances = nparray(binvariances)
binndets = nparray(binndets)
theta_top = npsum(binvariances*(binndets - 1)) / (npsum(binndets) -
goodbins)
theta_bot = npvar(pmags,ddof=1)
theta = theta_top/theta_bot
return theta
def pwd_phasebin(phases, mags, binsize=0.002, minbin=9):
'''
This bins the phased mag series using the given binsize.
'''
bins = np.arange(0.0, 1.0, binsize)
binnedphaseinds = npdigitize(phases, bins)
binnedphases, binnedmags = [], []
for x in npunique(binnedphaseinds):
thisbin_inds = binnedphaseinds == x
thisbin_phases = phases[thisbin_inds]
thisbin_mags = mags[thisbin_inds]
if thisbin_inds.size > minbin:
binnedphases.append(npmedian(thisbin_phases))
binnedmags.append(npmedian(thisbin_mags))
return np.array(binnedphases), np.array(binnedmags)
def aov_theta(times, mags, errs, frequency,
binsize=0.05, minbin=9):
'''Calculates the Schwarzenberg-Czerny AoV statistic at a test frequency.
Parameters
----------
times,mags,errs : np.array
The input time-series and associated errors.
frequency : float
The test frequency to calculate the theta statistic at.
binsize : float
The phase bin size to use.
minbin : int
The minimum number of items in a phase bin to consider in the
calculation of the statistic.
Returns
-------
theta_aov : float
The value of the AoV statistic at the specified `frequency`.
'''
period = 1.0/frequency
fold_time = times[0]
phased = phase_magseries(times,
mags,
period,
fold_time,
wrap=False,
sort=True)
phases = phased['phase']
pmags = phased['mags']
bins = nparange(0.0, 1.0, binsize)
ndets = phases.size
binnedphaseinds = npdigitize(phases, bins)
bin_s1_tops = []
bin_s2_tops = []
binndets = []
goodbins = 0
all_xbar = npmedian(pmags)
for x in npunique(binnedphaseinds):
thisbin_inds = binnedphaseinds == x
thisbin_mags = pmags[thisbin_inds]
if thisbin_mags.size > minbin:
thisbin_ndet = thisbin_mags.size
thisbin_xbar = npmedian(thisbin_mags)
# get s1
thisbin_s1_top = (
thisbin_ndet *
(thisbin_xbar - all_xbar) *
(thisbin_xbar - all_xbar)
)
# get s2
thisbin_s2_top = npsum((thisbin_mags - all_xbar) *
(thisbin_mags - all_xbar))
bin_s1_tops.append(thisbin_s1_top)
bin_s2_tops.append(thisbin_s2_top)
binndets.append(thisbin_ndet)
goodbins = goodbins + 1
# turn the quantities into arrays
bin_s1_tops = nparray(bin_s1_tops)
bin_s2_tops = nparray(bin_s2_tops)
binndets = nparray(binndets)
# calculate s1 first
s1 = npsum(bin_s1_tops)/(goodbins - 1.0)
# then calculate s2
s2 = npsum(bin_s2_tops)/(ndets - goodbins)
theta_aov = s1/s2
return theta_aov
def stellingwerf_pdm_theta(times,
mags,
errs,
frequency,
binsize=0.05,
minbin=9):
'''
This calculates the Stellingwerf PDM theta value at a test frequency.
Parameters
----------
times,mags,errs : np.array
The input time-series and associated errors.
frequency : float
The test frequency to calculate the theta statistic at.
binsize : float
The phase bin size to use.
minbin : int
The minimum number of items in a phase bin to consider in the
calculation of the statistic.
Returns
-------
theta_pdm : float
The value of the theta statistic at the specified `frequency`.
'''
period = 1.0 / frequency
fold_time = times[0]
phased = phase_magseries(times,
mags,
period,
fold_time,
wrap=False,
sort=True)
phases = phased['phase']
pmags = phased['mags']
bins = nparange(0.0, 1.0, binsize)
binnedphaseinds = npdigitize(phases, bins)
binvariances = []
binndets = []
goodbins = 0
for x in npunique(binnedphaseinds):
thisbin_inds = binnedphaseinds == x
thisbin_mags = pmags[thisbin_inds]
if thisbin_mags.size > minbin:
thisbin_variance = npvar(thisbin_mags, ddof=1)
binvariances.append(thisbin_variance)
binndets.append(thisbin_mags.size)
goodbins = goodbins + 1
# now calculate theta
binvariances = nparray(binvariances)
binndets = nparray(binndets)
theta_top = npsum(binvariances *
(binndets - 1)) / (npsum(binndets) - goodbins)
theta_bot = npvar(pmags, ddof=1)
theta = theta_top / theta_bot
return theta
def aov_theta(times, mags, errs, frequency, binsize=0.05, minbin=9):
'''Calculates the Schwarzenberg-Czerny AoV statistic at a test frequency.
'''
period = 1.0 / frequency
fold_time = times[0]
phased = phase_magseries(times,
mags,
period,
fold_time,
wrap=False,
sort=True)
phases = phased['phase']
pmags = phased['mags']
bins = np.arange(0.0, 1.0, binsize)
nbins = bins.size
ndets = phases.size
binnedphaseinds = npdigitize(phases, bins)
bin_s1_tops = []
bin_s2_tops = []
binndets = []
goodbins = 0
all_xbar = npmedian(pmags)
for x in npunique(binnedphaseinds):
thisbin_inds = binnedphaseinds == x
thisbin_phases = phases[thisbin_inds]
thisbin_mags = pmags[thisbin_inds]
if thisbin_mags.size > minbin:
thisbin_ndet = thisbin_mags.size
thisbin_xbar = npmedian(thisbin_mags)
# get s1
thisbin_s1_top = (thisbin_ndet * (thisbin_xbar - all_xbar) *
(thisbin_xbar - all_xbar))
# get s2
thisbin_s2_top = npsum(
(thisbin_mags - all_xbar) * (thisbin_mags - all_xbar))
bin_s1_tops.append(thisbin_s1_top)
bin_s2_tops.append(thisbin_s2_top)
binndets.append(thisbin_ndet)
goodbins = goodbins + 1
# turn the quantities into arrays
bin_s1_tops = nparray(bin_s1_tops)
bin_s2_tops = nparray(bin_s2_tops)
binndets = nparray(binndets)
# calculate s1 first
s1 = npsum(bin_s1_tops) / (goodbins - 1.0)
# then calculate s2
s2 = npsum(bin_s2_tops) / (ndets - goodbins)
theta_aov = s1 / s2
return theta_aov
def execArisel(y, w, pRegions, inits = 3, initialSolution = [],
convTabu = 0, tabuLength = 10):
"""Automatic Rationalization with Initial Seed Location
ARiSeL, proposed by [Duque_Church2004]_ , aggregates N areas into P
spatially contiguous regions while minimizing intra-regional heterogeneity
(measured as the within-cluster sum of squares from each area to the
attribute centroid of its cluster). This algorithm is a modification of
Openshaw's AZP-tabu [Openshaw_Rao1995]_. In ARISeL the construction of a
initial feasible solution is repeated several times (inits) before
running Tabu Search algorithm [Glover1977]_.
Duque and Church argue that:
- constructing and initial feasible solution is computationally less
expensive than performing local search.
- local search by moving bordering areas between region do not allow
an extensive search in the solution space and it is computationally
expensive.
Based on those two ideas, the authors propose to generate as many
different initial feasible solutions and run Tabu search on the best
initial solution obtained so far.
The initial solution follows a "growing regions" strategy. It starts with
a initial set of seeds (as many seed as regions) selected using the
K-means++ algorithm. From those seeds, other neighbouring areas are
assigned to its closest (in attribute space) growing region. This strategy
has proven better results. ::
Layer.cluster('arisel', vars, regions, <wType>, <std>, <inits>,
<initialSolution>, <convTabu>, <tabuLength>,
<dissolve>, <dataOperations>)
:keyword vars: Area attribute(s) (e.g. ['SAR1','SAR2'])
:type vars: list
:keyword regions: Number of regions
:type regions: integer
:keyword wType: Type of first-order contiguity-based spatial matrix: 'rook'
or 'queen'. Default value wType = 'rook'.
:type wType: string
:keyword std: If = 1, then the variables will be standardized.
:type std: binary
:keyword inits: number of initial feasible solutions to be constructed
before applying Tabu Search.
:type inits: integer. Default value inits = 5.
:keyword initialSolution: List with a initial solution vector. It is useful
when the user wants a solution that is not very different from a preexisting
solution (e.g. municipalities,districts, etc.). Note that the number of
regions will be the same as the number of regions in the initial feasible
solution (regardless the value you assign to parameter "regions").
IMPORTANT: make sure you are entering a feasible solution and according to
the W matrix you selected, otherwise the algorithm will not converge.
:type initialSolution: list
:keyword convTabu: Stop the search after convTabu nonimproving moves
(nonimproving moves are those moves that do not improve the current
solution.
Note that "improving moves" are different to "aspirational moves").
If convTabu=0 the algorithm will stop after Int(M/N) nonimproving moves.
Default value convTabu = 0.
:type convTabu: integer
:keyword tabuLength: Number of times a reverse move is prohibited. Default
value *tabuLength = 10*.
:type tabuLength: integer
:keyword dissolve: If = 1, then you will get a "child" instance of the layer
that contains the new regions. Default value *dissolve = 0*. **Note:**.
Each child layer is saved in the attribute *layer.results*. The first
algorithm that you run with *dissolve=1* will have a child layer in
*layer.results[0]*; the second algorithm that you run with *dissolve=1* will
be in *layer.results[1]*, and so on. You can export a child as a shapefile
with *layer.result[<1,2,3..>].exportArcData('filename')*
:type dissolve: binary
:keyword dataOperations: Dictionary which maps a variable to a list of
operations to run on it. The dissolved layer will contains in it's data all
the variables specified in this dictionary. Be sure to check the input
layer's fieldNames before use this utility.
:type dataOperations: dictionary
The dictionary structure must be as showed bellow.
>>> X = {}
>>> X[variableName1] = [function1, function2,....]
>>> X[variableName2] = [function1, function2,....]
Where functions are strings wich represents the name of the
functions to be used on the given variableName. Functions
could be,'sum','mean','min','max','meanDesv','stdDesv','med',
'mode','range','first','last','numberOfAreas. By deffault just
ID variable is added to the dissolved map.
"""
lenY = len(y)
start = 0.0
time2 = 0.0
print "Running original Arisel algorithm"
print "Number of areas: ", lenY
if initialSolution:
print "Number of regions: ", len(npunique(initialSolution))
pRegions = len(set(initialSolution))
else:
print "Number of regions: ", pRegions
if pRegions >= lenY:
message = "\n WARNING: You are aggregating "+str(lenY)+" into"+\
str(pRegions)+" regions!!. The number of regions must be an integer"+\
" number lower than the number of areas being aggregated"
raise Exception(message)
if convTabu <= 0:
convTabu = lenY/pRegions # convTabu = 230*numpy.sqrt(pRegions)
distanceType = "EuclideanSquared"
distanceStat = "Centroid"
objectiveFunctionType = "SS"
selectionType = "Minimum"
am = AreaManager(w, y, distanceType)
extendedMemory = ExtendedMemory()
pool = Pool(processes = cpu_count())
procs = []
start = tm.time()
for dummy in xrange(inits):
ans = pool.apply_async(constructPossible, [am, pRegions,
initialSolution,
distanceType,
distanceStat,
selectionType,
objectiveFunctionType])
procs.append(ans)
results = []
for p in procs:
results.append(p.get())
tmp_ans = extendedMemory
for rm in results:
if rm.objInfo < tmp_ans.objInfo:
tmp_ans = rm
rm = tmp_ans
extendedMemory.updateExtendedMemory(rm)
rm.recoverFromExtendedMemory(extendedMemory)
print "INITIAL SOLUTION: ", rm.returnRegions(), "\nINITIAL OF: ", rm.objInfo
rm.tabuMove(tabuLength=tabuLength, convTabu=convTabu)
time2 = tm.time() - start
Sol = rm.regions
Of = rm.objInfo
print "FINAL SOLUTION: ", Sol, "\nFINAL OF: ", Of
output = { "objectiveFunction": Of,
"runningTime": time2,
"algorithm": "arisel",
"regions": len(Sol),
"r2a": Sol,
"distanceType": distanceType,
"distanceStat": distanceStat,
"selectionType": selectionType,
"ObjectiveFuncionType": objectiveFunctionType}
return output
