def plot(self, hardcopy=None):
if hardcopy:
R.png(hardcopy, width=1024, height=768, type="cairo")
R.require('qvalue')
# build a qobj
R.assign("pval", self.mPValues)
R.assign("pi0", self.mPi0)
R.assign("qval", self.mQValues)
R.assign("lambda", self.mLambda)
R("""qobj <-list( pi0=pi0, qvalues=qval, pvalues=pval, lambda=lambda)"""
)
R(""" class(qobj) <- "qvalue" """)
R("""qplot(qobj)""")
if hardcopy:
R.dev_off()
def main():
parser = E.OptionParser( version = "%prog version: $Id: rates2rates.py 2781 2009-09-10 11:33:14Z andreas $", usage = globals()["__doc__"])
parser.add_option( "--output-filename-pattern", dest="output_filename_pattern", type="string",
help="pattern for additional output files [%default]." )
parser.add_option( "--input-filename-neutral", dest="input_filename_neutral", type="string",
help="a tab-separated file with rates and G+C content in neutrally evolving regions [%default]." )
parser.set_defaults(
input_filename_neutral = None,
output_filename_pattern = "%s",
normalize = True,
hardcopy = None,
)
(options, args) = E.Start( parser, add_csv_options = True )
if not options.input_filename_neutral:
raise ValueError( "please supply a file with neutral rates." )
lines = options.stdin.readlines()
if len(lines) == 0:
raise IOError ( "no input" )
from rpy import r as R
import rpy
R.png( options.output_filename_pattern % "fit" + ".png", width=1024, height=768, type="cairo")
matrix, headers = readRates( open( options.input_filename_neutral, "r" ) )
R.assign("matrix", matrix)
R.assign("headers", headers)
nref = R( """length( matrix[,1] )""" )
dat = R("""dat <- data.frame(x = matrix[,2], y = matrix[,3])""")
mod = R("""mod <- lm( y ~ x, dat)""")
R("""plot( matrix[,2], matrix[,3], cex=%s, col="blue", pch="o", xlab="%s", ylab="%s" %s)""" % (0.3, headers[1], headers[2], "") )
R("""new <- data.frame(x = seq( min(matrix[,2]), max(matrix[,2]), (max(matrix[,2]) - min(matrix[,2])) / 100))""")
R("""predict(mod, new, se.fit = TRUE)""")
R("""pred.w.plim <- predict(mod, new, interval="prediction")""")
R("""pred.w.clim <- predict(mod, new, interval="confidence")""")
R("""matpoints(new$x,cbind(pred.w.clim, pred.w.plim[,-1]), lty=c(1,2,2,3,3), type="l")""")
R.mtext(
"y = %f * x + %f, r=%6.4f, n=%i" % (mod["coefficients"]["x"],
mod["coefficients"]["(Intercept)"],
R("""cor( dat )[2]"""),
nref ),
3,
cex = 1.0)
R("""mean_rate <- mean( matrix[,3] )""")
data_matrix, data_headers = readRates( lines )
R.assign("data_matrix", data_matrix)
R.assign("data_headers", data_headers)
ndata = R( """length( data_matrix[,1] )""" )
R("""points( data_matrix[,2], data_matrix[,3], cex=%s, col="red", pch="o" %s)""" % (0.3, "") )
R("""topred <- data.frame( x = data_matrix[,2] )""")
R("""corrected_rates <- predict( mod, topred, se.fit = TRUE )""")
uncorrected = R("""uncorrected <- data_matrix[,3] / mean_rate """)
corrected = R("""corrected <- as.vector(data_matrix[,3] / corrected_rates$fit)""")
R.dev_off()
R.png( options.output_filename_pattern % "correction" + ".png", width=1024, height=768, type="cairo")
R("""plot( uncorrected, corrected, cex=%s, col="blue", pch="o", xlab="uncorrected rate", ylab="corrected rate" %s)""" % (0.3, "") )
R.dev_off()
E.Stop()
def plot(outfile, data, out_format='png'):
w = int(round(len(data) / 4.0))
if out_format == 'png':
r.png(outfile, width=w * 100, height=1000, res=72)
elif out_format == 'pdf':
r.pdf(outfile, width=w, height=10)
else:
raise Exception('Unrecognised format: ' + str(out_format))
print("total: " + str(len(data)))
series = []
points = {'translate': [], 'preprocessing': []}
for dat in data:
points['translate'].append(float(dat['translate']))
points['preprocessing'].append(float(dat['preprocessing']))
xlabels = []
for k, v in data[0].iteritems():
if k not in ["problem", 'translate', 'preprocessing']:
series.append(k)
points[k] = []
index = 0
for dat in data:
for k in series:
if dat[k] != 'no-plan':
points[k].append(float(dat[k]) + \
points['translate'][index] + \
points['preprocessing'][index])
else:
points[k].append(-1000)
xlabels.append(dat['problem'])
index += 1
max_value = max(iter([max(iter(points[k])) for k in series]))
yrange = (0, max_value)
legend_labels = []
x = [i for i in range(1, len(points['translate']) + 1)]
y = [-1000 for i in x]
r.par(mar=(7, 5, 4, 2))
r.plot(x,
y,
main='',
xlab="",
ylab='',
xaxt='n',
yaxt='n',
pch=0,
ylim=yrange,
mgp=(5, 1, 0))
r.mtext("Problem", side=1, line=5)
r.mtext("CPU Time (s)", side=2, line=3)
pch_start = 1
pch_index = pch_start
# plotting "translate"
#r.plot(x, points['translate'], main='',
# xlab='', ylab='Time (s)',
# xaxt='n', yaxt='n',
# pch=0, ylim=yrange)
#legend_labels.append('translate')
r.lines(x, points['translate'], lty=1)
# preprocessing -- Removed since it's insignificant
#r.points(x, points['preprocessing'], pch=pch_index)
#pch_index =+ 1
# planner output
for k in series:
if k != 'translate' and k != 'preporcessing':
r.points(x, points[k], pch=pch_index)
pch_index += 1
legend_labels.append("FD+" + k.upper())
# put x-axis labels
for i in range(0, len(xlabels)):
r.axis(side=1, at=i + 1, labels=xlabels[i], las=2)
# put y-axis labels
base, step = get_y_step(max_value)
print("base: " + str(base) + " -- step: " + str(step))
y = base
for i in range(0, step):
r.axis(side=2, at=y, labels=str(y), las=2)
y += base
# legend
r.legend(1,
max_value,
legend_labels,
pch=[i for i in range(pch_start, pch_index)])
r.dev_off()