def funcion(dato,opciones):
from rpy import r
diccionario={}
if opciones.has_key("Cuantiles"):
if opciones["Cuantiles"][u"DirecciónCola"]=='izquierda':
sentido=True
else:
sentido=False
diccionario["cuantiles"]=r.qnorm([float(opciones["Cuantiles"]["Probabilidad"])],mean=float(opciones["Cuantiles"]["Media"]),sd=float(opciones["Cuantiles"][u"Desviación"]),lower_tail=sentido)
if opciones.has_key("Probabilidades"):
if opciones["Probabilidades"][u"DirecciónCola"]=='izquierda':
sentido=True
else:
sentido=False
diccionario["probabilidades"]=r.pnorm([float(opciones["Probabilidades"]["Valores"])],mean=float(opciones["Probabilidades"]["Media"]),sd=float(opciones["Probabilidades"][u"Desviación"]),lower_tail=sentido)
if opciones.has_key(u"Gráfica"):
import random
nombrefichero="/tmp/driza"+str(random.randint(1,99999))+".png"
diccionario["ruta"]=nombrefichero
r.png(nombrefichero) #Directorio temporal de la config
lista=r.seq(-3.291, 3.291, length=100)
if opciones[u"Gráfica"]["Tipografica"]=="Densidad":
etiquetay="Densidad"
mifuncion=r.dnorm
else:
etiquetay="Probabilidad acumulada"
mifuncion=r.pnorm
r.plot(lista, mifuncion(lista, mean=float(opciones[u"Gráfica"]["Media"]), sd=float(opciones[u"Gráfica"][u"Desviación"])), xlab="x", ylab=etiquetay, main=r.expression(r.paste("Normal Distribution: ", "mu", " = 0, ", "sigma", " = 1")), type="l")
r.abline(h=0, col="gray")
r.dev_off()
return diccionario
def qqplot_density(samples, density_x, density_y):
from rpy import r
# LAME: should do better quantile calculation. r.quantile() returns a
# hard-to-use dictionary unfortunately. r.approx() as per ?quantile, maybe.
percs = list(np.arange(1,100, .1))
data_perc=np.percentile(samples, percs)
real_perc=np.percentile(r.sample(density_x, len(samples)*10, prob=density_y, replace=True), percs)
r.plot(real_perc, data_perc, xlab='',ylab='',main='');
r.abline(a=0,b=1,col='blue')
def trace_plots(history, burnin):
from rpy import r
h = np.array(history)
print "{} total, {} burning, {} remaining".format(len(history), burnin, len(history)-burnin)
r.par(mfrow=[2,2])
r.acf(h[burnin:])
r.plot(h,xlab='',ylab='',main='')
r.abline(v=burnin, col='blue')
#r.hist(h[burnin:],breaks=30,xlab='',ylab='',main='histogram')
r.plot(r.density(h[burnin:]), xlab='',ylab='',main='density')
def qqplot_density(samples, density_x, density_y):
from rpy import r
# LAME: should do better quantile calculation. r.quantile() returns a
# hard-to-use dictionary unfortunately. r.approx() as per ?quantile, maybe.
percs = list(np.arange(1, 100, .1))
data_perc = np.percentile(samples, percs)
real_perc = np.percentile(
r.sample(density_x, len(samples) * 10, prob=density_y, replace=True),
percs)
r.plot(real_perc, data_perc, xlab='', ylab='', main='')
r.abline(a=0, b=1, col='blue')
def trace_plots(history, burnin):
from rpy import r
h = np.array(history)
print "{} total, {} burning, {} remaining".format(len(history), burnin,
len(history) - burnin)
r.par(mfrow=[2, 2])
r.acf(h[burnin:])
r.plot(h, xlab='', ylab='', main='')
r.abline(v=burnin, col='blue')
#r.hist(h[burnin:],breaks=30,xlab='',ylab='',main='histogram')
r.plot(r.density(h[burnin:]), xlab='', ylab='', main='density')
def plots(regression_o, getData_o):
"""Plots the dataset with a regression line and a boxplot using R."""
fname1 = 'car_regress.pdf'
r.pdf(fname1)
r.plot(getData_o, ylab='dist', xlab='speed')
r.abline(regression_o['(Intercept)'], regression_o['y'], col='red')
r.dev_off()
fname2 = 'car_hist.pdf'
r.pdf(fname2)
r.boxplot(getData_o, names=['dist', 'speed'])
r.dev_off()
return fname1, fname2
def plotArray(self, hitDataParam, filename, plotDir=None, labeledPosD=None,
title='', type='png',
label_colors = True, colVecL=None, legend_plot = False, legend_filename=None,
control_color='black', numeric_as_int = False,
grid=True, max_val=0.3, min_val=0.0, cut_data = True):
if plotDir is None:
plotDir = self.plotDir
full_filename = os.path.join(plotDir, filename)
#hitData = copy.deepcopy(hitDataParam)
hitData = hitDataParam
#nrow = len(hitData)
#ncol = len(hitData[0])
nrow = self.nb_row
ncol = self.nb_col
xL = range(1, self.nb_col + 1)
yL = range(1, self.nb_row + 1)
rdev = plot_utilities.RDevice(name = full_filename, title=title, plotType=type,
width=640, height=250)
if label_colors:
# in this case, the data is considered to consist of labels
labelL = []
for i in range(nrow):
for j in range(ncol):
if hitData[i][j] not in labelL:
labelL.append(hitData[i][j])
if colVecL is None:
colVecL = r.rainbow(len(labelL))
colBreaksL = range(len(colVecL) + 1)
if legend_plot:
self.plotLabelLegend(colVecL, plotType=type, filename=legend_filename)
else:
# in this case, the data is numeric.
if cut_data:
max_rel_cell_count = max_val
min_rel_cell_count = 0.0
for i in range(nrow):
for j in range(ncol):
hitData[i][j] = max(min_rel_cell_count, min(max_rel_cell_count, hitData[i][j]))
else:
max_rel_cell_count = max([max(x) for x in hitData.tolist() ])
min_rel_cell_count = min([min(x) for x in hitData.tolist() ])
if numeric_as_int:
nb_colors = max_rel_cell_count
else:
nb_colors = 500
if colVecL is None:
pattern = [(0,0,0),(0.7,0,0),(1,1,0),(1,1,1)]
colVecL = colors.make_colors(pattern, nb_colors)
colBreaksL = [1.0/ (len(colVecL) - 1) * x * (max_rel_cell_count - min_rel_cell_count) + min_rel_cell_count for x in range(len(colVecL) + 1)]
if legend_plot:
self.plotNumLegend(colVecL, colBreaksL, 16, filename=legend_filename, type=type, int_labels=numeric_as_int, legendDir = plotDir)
axisSize = .8
r("par(mar=c(1.6,1.6,0.1,0.1))")
r.image(xL, yL, r.t(hitData), axes = False, ann=False, cex=1, col=colVecL, breaks=colBreaksL)
r.box()
if not labeledPosD is None:
for label in labeledPosD.keys():
posL = labeledPosD[label]
if len(posL) > 0:
xlL = [(int(x)-1) % self.nb_col + 1 for x in posL]
ylL = [(int(x)-1) / self.nb_col + 1 for x in posL]
r.points(xlL, ylL, pch=label, col=control_color, cex=axisSize)
print
print xlL
print ylL
print
# grid
if grid:
for i in range(self.nb_col):
r.abline(h=i+.5, lty=3, lwd=1, col='grey')
for i in range(self.nb_row):
r.abline(v=i+.5, lty=3, lwd=1, col='grey')
r.axis(1, at=xL, labels=[str(x) for x in xL], tick=False, line=-1.0, cex_axis=axisSize)
r.axis(2, at=yL, labels=[str(y) for y in yL], tick=False, line=-1.0, cex_axis=axisSize)
rdev.close()
return
from rpy import r
my_x = [5.05, 6.75, 3.21, 2.66]
my_y = [1.65, 26.5, -5.93, 7.96]
ls_fit = r.lsfit(my_x,my_y)
gradient = ls_fit['coefficients']['X']
yintercept= ls_fit['coefficients']['Intercept']
r.png("scatter_regression.png", width=400, height=350)
r.plot(x=my_x, y=my_y, xlab="x", ylab="y", xlim=(0,7), ylim=(-16,27),
main="Example Scatter with regression")
r.abline(a=yintercept, b=gradient, col="red")
r.dev_off()
def plotVertical(self, lty=3, col='lightgray', **other):
from rpy import r
for kmag in self.kmags():
r.abline(v=kmag, lty=lty, col=col, **other)
#thisRun.sort()
print "top modes:"
print "%6s %6s %7s %9s"%('kmag', 'Sk', 'kmag/L', 'L/kmag')
for Sk, kmag in zip(v_sk, v_kmag)[-10:]:
print "%6.3f %6.3f %7.5f %9.5f"%(kmag, Sk, kmag/L, L/kmag)
r.plot(v_kmag, v_sk,
xlab="", ylab="", type="l",
ylim=(0., 15)
#ylim=(0., 15000)
)
# plots a line at each k-vector point.
for kmag in v_kmag:
r.abline(v=kmag, lty=3, col="lightgray")
# plots each individual k-vector
for kmag, SkArray in zip(v_kmag, StructCorr.SkArrays()):
r.points(x=[kmag]*len(SkArray), y=SkArray,
col="blue", pch="x")
# plots the standard deviation of the by-kvector lists.
for kmag, SkArray in zip(v_kmag, StructCorr.SkArrays()):
r.points(x=kmag, y=numpy.std(SkArray),
col="red", pch="s")
# plots it using the average of all k-vectors -- this should line up
# exactly
#for kmag, SkArray in zip(v_kmag, StructCorr.SkArrays()):
# r.points(x=kmag, y=numpy.average(SkArray),
# col="green", pch="x")
