def smartReadMat(filename):
"""Read a matrix from a file, determining the file type automatically.
This function returns a pair (matrix,fieldnames). The filetype is
determined automatically from the extension. The following formats are
currently supported:
- '.amat' : PLearn ascii matrix format
- '.pmat' : PLearn binary matrix format
- '.csv' : Text-file comma-separated values format
"""
if filename.endswith(".amat"):
arr, fieldnames = readAMat(filename)
elif filename.endswith(".pmat"):
pmat = PMat(filename)
arr = pmat.getRows(0, pmat.length)
fieldnames = pmat.fieldnames
pmat.close()
elif filename.endswith(".csv"):
# Use CSV sniffer to detect presence of header.
sniffer = csv.Sniffer()
f = open(filename)
sample = f.read(1000)
has_header = sniffer.has_header(sample)
f.seek(0)
# Load csv into array
csv_reader = csv.reader(f)
if has_header:
fieldnames = csv_reader.next()
arr = numpy.numarray.array([[float(value) for value in fields] for fields in csv_reader])
if not has_header:
# Generate fake fieldnames
fieldnames = ['field%d' % (i + 1) for i in range(arr.shape[1])]
f.close()
else:
raise ValueError, "Unrecognized file type for '%s'; valid extensions are: " \
"{'.amat', '.pmat', '.csv'}" % filename
return arr, fieldnames
m = array([[stats[k][i] for i in range(self.width())] for k in sk])
_printMatrix(m, sk, self.fieldnames, os, pretty)
print "\nCovariance Matrix:"
_printMatrix(stats["COV"], self.fieldnames, self.fieldnames, os, pretty)
print "\nCorrelation Matrix:"
_printMatrix(stats["CORR"], self.fieldnames, self.fieldnames, os, pretty)
if __name__ == "__main__":
from plearn.vmat.readAMat import readAMat
ut,fieldnames = readAMat(
os.path.join(
ppath.ppath('PLEARNDIR'),
'examples', 'data', 'test_suite',
'top_100_test.amat'
)
)
sc = StatsCollector(fieldnames)
sc.update(ut)
sc.printStats(sys.stdout, False)
print "\nAfter accumulating some more:"
sc.update(ut)
sc.printStats(sys.stdout, False)
print "\nAfter forgetting:"
sc.forget(fieldnames)
sc.printStats(sys.stdout, False)
"N", "NMISSING", "NNONMISSING", "E", "V", "STDDEV", "STDERR",
"SUM", "SUMSQ", "MIN", "ARGMIN", "MAX", "ARGMAX"
]
m = array([[stats[k][i] for i in range(self.width())] for k in sk])
_printMatrix(m, sk, self.fieldnames, os, pretty)
print "\nCovariance Matrix:"
_printMatrix(stats["COV"], self.fieldnames, self.fieldnames, os,
pretty)
print "\nCorrelation Matrix:"
_printMatrix(stats["CORR"], self.fieldnames, self.fieldnames, os,
pretty)
if __name__ == "__main__":
from plearn.vmat.readAMat import readAMat
ut, fieldnames = readAMat(
os.path.join(ppath.ppath('PLEARNDIR'), 'examples', 'data',
'test_suite', 'top_100_test.amat'))
sc = StatsCollector(fieldnames)
sc.update(ut)
sc.printStats(sys.stdout, False)
print "\nAfter accumulating some more:"
sc.update(ut)
sc.printStats(sys.stdout, False)
print "\nAfter forgetting:"
sc.forget(fieldnames)
sc.printStats(sys.stdout, False)
