def _write_nab_file(self,fd,values,formats,names,masktype=None):
bfh = BinaryFileHeader(formats=formats,names=names,masktype=masktype)
fd.write(bfh.make_header()+'\n')
if not len(values):
return
try:
tmpdata = narec.array(values,formats=formats,names=names)
except MemoryError:
log.warning('got MemoryError, trying alternate write method')
tmpdata = None
# TANYA: The only error that numpy.core.records.array raises is ValueError; commenting out the below...
#except self.BufferError:
# log.info('reached memmap limit on system, using alternate write method')
# tmpdata = None
if tmpdata is not None:
tmpdata.tofile(fd)
del tmpdata
return
# write using small groups to conserve memory
total = len(values[0])
grpsize = max(1,min(1024*1024,total/16))
for i in range(0,total,grpsize):
valslice = [arr[i:i+grpsize] for arr in values]
tmpdata = narec.array(valslice,formats=formats,names=names)
tmpdata.tofile(fd)
del tmpdata
def to_array(records: Iterator[Purchase]) -> recarray:
return rec.array(list(records),
dtype=dtype([('report', unicode(8)), ('date', date64),
('report_date', date64), ('seller', uint8),
('arrangement', uint8), ('basis', uint8),
('head_count', uint32),
('avg_price', float32),
('low_price', float32),
('high_price', float32)]))
def to_array(records: Iterator[Sales]) -> recarray:
return rec.array(list(records), dtype=dtype([
('report', unicode(8)),
('date', date64),
('report_date', date64),
('type', uint8),
('description', unicode(64)),
('weight', uint32),
('avg_price', float32),
('low_price', float32),
('high_price', float32)
]))
def to_array(records: Iterator[Cutout]) -> recarray:
return rec.array(list(records),
dtype=dtype([('report', unicode(8)), ('date', date64),
('report_date', date64),
('primal_loads', float32),
('trimming_loads', float32),
('carcass_price', float32),
('loin_price', float32),
('butt_price', float32),
('picnic_price', float32),
('rib_price', float32),
('ham_price', float32),
('belly_price', float32)]))
def columnsOfType(self, columnames, rows=None, colTypes=None):
if colTypes == None:
return self.columns(columnames, rows)
dataSubset = self.columns(columnames, rows)
dataCols = []
dtypeInput = []
for i in range(len(columnames)):
colName = columnames[i]
colType = colTypes[colName]
dataCols.append(dataSubset.data[:,i].astype(colType))
dataSubset.data = rec.array(dataCols)
return dataSubset
def to_array(records: Iterator[Slaughter]) -> recarray:
return rec.array(list(records),
dtype=dtype([('report', unicode(8)), ('date', date64),
('report_date', date64), ('seller', uint8),
('arrangement', uint8), ('basis', uint8),
('head_count', uint32),
('base_price', float32),
('net_price', float32),
('low_price', float32),
('high_price', float32),
('live_weight', float32),
('carcass_weight', float32),
('sort_loss', float32), ('backfat', float32),
('loin_depth', float32),
('loineye_area', float32),
('lean_percent', float32)]))
from statsmodels.tools.testing import Holder
var_results = Holder()
var_results.comment = 'VAR test data converted from vars_results.npz'
var_results.causality = array([
(9.317172089406967e-08,),
(0.5183914225971917,),
(4.8960835385969403e-14,)],
dtype=[('causedby', 'float')])
var_results.name = 'var_results'
var_results.orthirf = array({
'realgdp': rec.array([
(0.007557357219752236, 0.003948403413668315, 0.02972434157321242),
(0.0015408726821578582, 0.0010664916255201816, 0.00923575489996933),
(0.0015874964105555918, 0.0010551760558416706, 0.006102514196485799),
(0.0007262051539604352, 0.0005562787500837443, 0.003199064883156089),
(0.0005537000868358786, 0.0003520396722562061, 0.0024372344590635623),
(0.0003079984190444812, 0.00021674897409108682, .0013369479853037147)],
dtype=[('realgdp', 'float'),
('realcons', 'float'),
('realinv', 'float')]),
'realinv': rec.array([
(.0, 0.0, 0.020741992721114832),
(.0006890376065674764, .0005338724781743238, 0.004676882806534488),
(.00017134455810606506, .000682084896451223, -0.0005205835547221123),
(.0005217378718553543, .00030179909990059973, 0.0026650577026759623),
(.00034979575853114173, .00022249591743758265, 0.0015804716569096742),
(.00017738402507880077, .00013384975583249413, 0.0007585745605878197)],
dtype=[('realgdp', 'float'),
('realcons', 'float'),
('realinv', 'float')]),
'realcons': rec.array([
var_results = Holder()
var_results.comment = 'VAR test data converted from vars_results.npz'
var_results.causality = array([
(9.317172089406967e-08,),
(0.5183914225971917,),
(4.8960835385969403e-14,)],
dtype=[('causedby', 'float')])
var_results.name = 'var_results'
var_results.orthirf = array({
'realgdp': rec.array([
(0.007557357219752236, 0.003948403413668315, 0.02972434157321242),
(0.0015408726821578582, 0.0010664916255201816, 0.00923575489996933),
(0.0015874964105555918, 0.0010551760558416706, 0.006102514196485799),
(0.0007262051539604352, 0.0005562787500837443, 0.003199064883156089),
(0.0005537000868358786, 0.0003520396722562061, 0.0024372344590635623),
(0.0003079984190444812, 0.00021674897409108682, .0013369479853037147)],
dtype=[('realgdp', 'float'),
('realcons', 'float'),
('realinv', 'float')]),
'realinv': rec.array([
(.0, 0.0, 0.020741992721114832),
(.0006890376065674764, .0005338724781743238, 0.004676882806534488),
(.00017134455810606506, .000682084896451223, -0.0005205835547221123),
(.0005217378718553543, .00030179909990059973, 0.0026650577026759623),
(.00034979575853114173, .00022249591743758265, 0.0015804716569096742),
(.00017738402507880077, .00013384975583249413, 0.0007585745605878197)],
dtype=[('realgdp', 'float'),
('realcons', 'float'),
('realinv', 'float')]),
'realcons': rec.array([
