def main(dictAlg, sortedAlgs, isBiobjective, outputdir='.', verbose=True, function_targets_line=True): # [1, 13, 101]
"""Generate one table per func with results of multiple algorithms."""
"""Difference with the first version:
* numbers aligned using the decimal separator
* premices for dispersion measure
* significance test against best algorithm
* table width...
Takes ``targetsOfInterest`` from this file as "input argument" to compute
the desired target values. ``targetsOfInterest`` might be configured via
config.
"""
# TODO: method is long, terrible to read, split if possible
bestalgentries = bestalg.loadBestAlgorithm(isBiobjective)
# Sort data per dimension and function
dictData = {}
dsListperAlg = list(dictAlg[i] for i in sortedAlgs)
for n, entries in enumerate(dsListperAlg):
tmpdictdim = entries.dictByDim()
for d in tmpdictdim:
tmpdictfun = tmpdictdim[d].dictByFunc()
for f in tmpdictfun:
dictData.setdefault((d, f), {})[n] = tmpdictfun[f]
nbtests = len(dictData)
funInfos = ppfigparam.read_fun_infos(isBiobjective)
for df in dictData:
# Generate one table per df
# first update targets for each dimension-function pair if needed:
targets = targetsOfInterest((df[1], df[0]))
targetf = targets[-1]
# best 2009
refalgentry = bestalgentries[df]
refalgert = refalgentry.detERT(targets)
refalgevals = (refalgentry.detEvals((targetf, ))[0][0])
refalgnbruns = len(refalgevals)
refalgnbsucc = numpy.sum(numpy.isnan(refalgevals) == False)
# Process the data
# The following variables will be lists of elements each corresponding
# to an algorithm
algnames = []
#algdata = []
algerts = []
algevals = []
algdisp = []
algnbsucc = []
algnbruns = []
algmedmaxevals = []
algmedfinalfunvals = []
algtestres = []
algentries = []
for n in sorted(dictData[df].keys()):
entries = dictData[df][n]
# the number of datasets for a given dimension and function (df)
# should be strictly 1. TODO: find a way to warn
# TODO: do this checking before... why wasn't it triggered by ppperprof?
if len(entries) > 1:
print entries
txt = ("There is more than a single entry associated with "
"folder %s on %d-D f%d." % (sortedAlgs[n], df[0], df[1]))
raise Exception(txt)
entry = entries[0]
algentries.append(entry)
algnames.append(sortedAlgs[n])
evals = entry.detEvals(targets)
#tmpdata = []
tmpdisp = []
tmpert = []
for i, e in enumerate(evals):
succ = (numpy.isnan(e) == False)
ec = e.copy() # note: here was the previous bug (changes made in e also appeared in evals !)
ec[succ == False] = entry.maxevals[succ == False]
ert = toolsstats.sp(ec, issuccessful=succ)[0]
#tmpdata.append(ert/refalgert[i])
if succ.any():
tmp = toolsstats.drawSP(ec[succ], entry.maxevals[succ == False],
[10, 50, 90], samplesize=samplesize)[0]
tmpdisp.append((tmp[-1] - tmp[0])/2.)
else:
tmpdisp.append(numpy.nan)
tmpert.append(ert)
algerts.append(tmpert)
algevals.append(evals)
#algdata.append(tmpdata)
algdisp.append(tmpdisp)
algmedmaxevals.append(numpy.median(entry.maxevals))
algmedfinalfunvals.append(numpy.median(entry.finalfunvals))
#algmedmaxevals.append(numpy.median(entry.maxevals)/df[0])
#algmedfinalfunvals.append(numpy.median(entry.finalfunvals))
algtestres.append(significancetest(refalgentry, entry, targets))
# determine success probability for Df = 1e-8
e = entry.detEvals((targetf ,))[0]
algnbsucc.append(numpy.sum(numpy.isnan(e) == False))
algnbruns.append(len(e))
# Process over all data
# find best values...
nalgs = len(dictData[df])
maxRank = 1 + numpy.floor(0.14 * nalgs) # number of algs to be displayed in bold
isBoldArray = [] # Point out the best values
algfinaldata = [] # Store median function values/median number of function evaluations
tmptop = getTopIndicesOfColumns(algerts, maxRank=maxRank)
for i, erts in enumerate(algerts):
tmp = []
for j, ert in enumerate(erts): # algi targetj
tmp.append(i in tmptop[j] or (nalgs > 7 and algerts[i][j] <= 3. * refalgert[j]))
isBoldArray.append(tmp)
algfinaldata.append((algmedfinalfunvals[i], algmedmaxevals[i]))
# significance test of best given algorithm against all others
best_alg_idx = numpy.array(algerts).argsort(0)[0, :] # indexed by target index
significance_versus_others = significance_all_best_vs_other(algentries, targets, best_alg_idx)[0]
# Create the table
table = []
tableHtml = []
spec = r'@{}c@{}|*{%d}{@{\,}r@{}X@{\,}}|@{}r@{}@{}l@{}' % (len(targets)) # in case StrLeft not working: replaced c@{} with l@{ }
spec = r'@{}c@{}|*{%d}{@{}r@{}X@{}}|@{}r@{}@{}l@{}' % (len(targets)) # in case StrLeft not working: replaced c@{} with l@{ }
extraeol = []
# Generate header lines
if with_table_heading:
header = funInfos[df[1]] if df[1] in funInfos.keys() else 'f%d' % df[1]
table.append([r'\multicolumn{%d}{@{\,}c@{\,}}{{\textbf{%s}}}'
% (2 * len(targets) + 2, header)])
extraeol.append('')
if function_targets_line is True or (function_targets_line and df[1] in function_targets_line):
if isinstance(targetsOfInterest, pproc.RunlengthBasedTargetValues):
curline = [r'\#FEs/D']
curlineHtml = ['<thead>\n<tr>\n<th>#FEs/D<br>REPLACEH</th>\n']
counter = 1
for i in targetsOfInterest.labels():
curline.append(r'\multicolumn{2}{@{}c@{}}{%s}' % i)
curlineHtml.append('<td>%s<br>REPLACE%d</td>\n' % (i, counter))
counter += 1
else:
curline = [r'$\Delta f_\mathrm{opt}$']
curlineHtml = ['<thead>\n<tr>\n<th>Δ f<sub>opt</sub><br>REPLACEH</th>\n']
counter = 1
for t in targets:
curline.append(r'\multicolumn{2}{@{\,}X@{\,}}{%s}'
% writeFEvals2(t, precision=1, isscientific=True))
curlineHtml.append('<td>%s<br>REPLACE%d</td>\n' % (writeFEvals2(t, precision=1, isscientific=True), counter))
counter += 1
# curline.append(r'\multicolumn{2}{@{\,}X@{}|}{%s}'
# % writeFEvals2(targets[-1], precision=1, isscientific=True))
curline.append(r'\multicolumn{2}{@{}l@{}}{\#succ}')
curlineHtml.append('<td>#succ<br>REPLACEF</td>\n</tr>\n</thead>\n')
table.append(curline)
extraeol.append(r'\hline')
# extraeol.append(r'\hline\arrayrulecolor{tableShade}')
curline = [r'ERT$_{\text{best}}$'] if with_table_heading else [r'\textbf{f%d}' % df[1]]
replaceValue = 'ERT<sub>best</sub>' if with_table_heading else ('<b>f%d</b>' % df[1])
curlineHtml = [item.replace('REPLACEH', replaceValue) for item in curlineHtml]
if isinstance(targetsOfInterest, pproc.RunlengthBasedTargetValues):
# write ftarget:fevals
counter = 1
for i in xrange(len(refalgert[:-1])):
temp="%.1e" %targetsOfInterest((df[1], df[0]))[i]
if temp[-2]=="0":
temp=temp[:-2]+temp[-1]
curline.append(r'\multicolumn{2}{@{}c@{}}{\textit{%s}:%s \quad}'
% (temp, writeFEvalsMaxPrec(refalgert[i], 2)))
replaceValue = '<i>%s</i>:%s' % (temp, writeFEvalsMaxPrec(refalgert[i], 2))
curlineHtml = [item.replace('REPLACE%d' % counter, replaceValue) for item in curlineHtml]
counter += 1
temp="%.1e" %targetsOfInterest((df[1], df[0]))[-1]
if temp[-2]=="0":
temp=temp[:-2]+temp[-1]
curline.append(r'\multicolumn{2}{@{}c@{}|}{\textit{%s}:%s }'
% (temp ,writeFEvalsMaxPrec(refalgert[-1], 2)))
replaceValue = '<i>%s</i>:%s' % (temp, writeFEvalsMaxPrec(refalgert[-1], 2))
curlineHtml = [item.replace('REPLACE%d' % counter, replaceValue) for item in curlineHtml]
else:
# write #fevals of the reference alg
counter = 1
for i in refalgert[:-1]:
curline.append(r'\multicolumn{2}{@{}c@{}}{%s \quad}'
% writeFEvalsMaxPrec(i, 2))
curlineHtml = [item.replace('REPLACE%d' % counter, writeFEvalsMaxPrec(i, 2)) for item in curlineHtml]
counter += 1
curline.append(r'\multicolumn{2}{@{}c@{}|}{%s}'
% writeFEvalsMaxPrec(refalgert[-1], 2))
curlineHtml = [item.replace('REPLACE%d' % counter, writeFEvalsMaxPrec(refalgert[-1], 2)) for item in curlineHtml]
# write the success ratio for the reference alg
tmp2 = numpy.sum(numpy.isnan(refalgevals) == False) # count the nb of success
curline.append('%d' % (tmp2))
if tmp2 > 0:
curline.append('/%d' % len(refalgevals))
replaceValue = '%d/%d' % (tmp2, len(refalgevals))
else:
replaceValue = '%d' % tmp2
curlineHtml = [item.replace('REPLACEF', replaceValue) for item in curlineHtml]
table.append(curline[:])
tableHtml.extend(curlineHtml[:])
tableHtml.append('<tbody>\n')
extraeol.append('')
#for i, gna in enumerate(zip((1, 2, 3), ('bla', 'blo', 'bli'))):
#print i, gna, gno
#set_trace()
# Format data
#if df == (5, 17):
#set_trace()
header = r'\providecommand{\ntables}{7}'
for i, alg in enumerate(algnames):
tableHtml.append('<tr>\n')
#algname, entries, irs, line, line2, succ, runs, testres1alg in zip(algnames,
#data, dispersion, isBoldArray, isItalArray, nbsucc, nbruns, testres):
commandname = r'\alg%stables' % numtotext(i)
# header += r'\providecommand{%s}{{%s}{}}' % (commandname, str_to_latex(strip_pathname(alg)))
header += r'\providecommand{%s}{\StrLeft{%s}{\ntables}}' % (commandname, str_to_latex(strip_pathname1(alg)))
curline = [commandname + r'\hspace*{\fill}'] # each list element becomes a &-separated table entry?
curlineHtml = ['<th>%s</th>\n' % str_to_latex(strip_pathname1(alg))]
for j, tmp in enumerate(zip(algerts[i], algdisp[i], # j is target index
isBoldArray[i], algtestres[i])):
ert, dispersion, isBold, testres = tmp
alignment = '@{\,}X@{\,}'
if j == len(algerts[i]) - 1:
alignment = '@{\,}X@{\,}|'
data = ert/refalgert[j]
# write star for significance against all other algorithms
str_significance_subsup = ''
str_significance_subsup_html = ''
if (len(best_alg_idx) > 0 and len(significance_versus_others) > 0 and
i == best_alg_idx[j] and nbtests * significance_versus_others[j][1] < 0.05):
logp = -numpy.ceil(numpy.log10(nbtests * significance_versus_others[j][1]))
logp = numpy.min((9, logp)) # not messing up the format and handling inf
str_significance_subsup = r"^{%s%s}" % (significance_vs_others_symbol, str(int(logp)) if logp > 1 else '')
str_significance_subsup_html = '<sup>%s%s</sup>' % (significance_vs_others_symbol_html, str(int(logp)) if logp > 1 else '')
# moved out of the above else: this was a bug!?
z, p = testres
if (nbtests * p) < 0.05 and data < 1. and z < 0.:
if not numpy.isinf(refalgert[j]):
tmpevals = algevals[i][j].copy()
tmpevals[numpy.isnan(tmpevals)] = algentries[i].maxevals[numpy.isnan(tmpevals)]
bestevals = refalgentry.detEvals(targets)
bestevals, bestalgalg = (bestevals[0][0], bestevals[1][0])
bestevals[numpy.isnan(bestevals)] = refalgentry.maxevals[bestalgalg][numpy.isnan(bestevals)]
tmpevals = numpy.array(sorted(tmpevals))[0:min(len(tmpevals), len(bestevals))]
bestevals = numpy.array(sorted(bestevals))[0:min(len(tmpevals), len(bestevals))]
#The conditions are now that ERT < ERT_best and
# all(sorted(FEvals_best) > sorted(FEvals_current)).
if numpy.isinf(refalgert[j]) or all(tmpevals < bestevals):
nbstars = -numpy.ceil(numpy.log10(nbtests * p))
# tmp2[-1] += r'$^{%s}$' % superscript
str_significance_subsup += r'_{%s%s}' % (significance_vs_ref_symbol,
str(int(nbstars)) if nbstars > 1 else '')
str_significance_subsup_html = '<sub>%s%s</sub>' % (significance_vs_ref_symbol_html,
str(int(nbstars)) if nbstars > 1 else '')
if str_significance_subsup:
str_significance_subsup = '$%s$' % str_significance_subsup
# format number in variable data
if numpy.isnan(data):
curline.append(r'\multicolumn{2}{%s}{.}' % alignment)
else:
if numpy.isinf(refalgert[j]):
curline.append(r'\multicolumn{2}{%s}{\textbf{%s}\mbox{\tiny (%s)}%s}'
% (alignment,
writeFEvalsMaxPrec(algerts[i][j], 2),
writeFEvalsMaxPrec(dispersion, precdispersion),
str_significance_subsup))
curlineHtml.append('<td sorttable_customkey=\"%f\"><b>%s</b> (%s)%s</td>\n'
% (algerts[i][j],
writeFEvalsMaxPrec(algerts[i][j], 2),
writeFEvalsMaxPrec(dispersion, precdispersion),
str_significance_subsup_html))
continue
tmp = writeFEvalsMaxPrec(data, precfloat, maxfloatrepr=maxfloatrepr)
tmpHtml = writeFEvalsMaxPrec(data, precfloat, maxfloatrepr=maxfloatrepr)
sortKey = data
if data >= maxfloatrepr or data < 0.01: # either inf or scientific notation
if numpy.isinf(data) and j == len(algerts[i]) - 1:
tmp += r'\,\textit{%s}' % writeFEvalsMaxPrec(algfinaldata[i][1], 0, maxfloatrepr=maxfloatrepr)
tmpHtml += '<i>%s</i>' % writeFEvalsMaxPrec(algfinaldata[i][1], 0, maxfloatrepr=maxfloatrepr)
sortKey = algfinaldata[i][1]
else:
tmp = writeFEvalsMaxPrec(data, precscien, maxfloatrepr=data)
if isBold:
tmpHtml = '<b>%s</b>' % tmp
tmp = r'\textbf{%s}' % tmp
if not numpy.isnan(dispersion):
tmpdisp = dispersion/refalgert[j]
if tmpdisp >= maxfloatrepr or tmpdisp < 0.005: # TODO: hack
tmpdisp = writeFEvalsMaxPrec(tmpdisp, precdispersion, maxfloatrepr=tmpdisp)
else:
tmpdisp = writeFEvalsMaxPrec(tmpdisp, precdispersion, maxfloatrepr=maxfloatrepr)
tmp += r'\mbox{\tiny (%s)}' % tmpdisp
tmpHtml += ' (%s)' % tmpdisp
curline.append(r'\multicolumn{2}{%s}{%s%s}' % (alignment, tmp, str_significance_subsup))
tmpHtml = tmpHtml.replace('$\infty$', '∞')
if (numpy.isinf(sortKey)):
sortKey = sys.maxint
curlineHtml.append('<td sorttable_customkey=\"%f\">%s%s</td>' % (sortKey, tmpHtml, str_significance_subsup_html))
else:
tmp2 = tmp.split('.', 1)
if len(tmp2) < 2:
tmp2.append('')
else:
tmp2[-1] = '.' + tmp2[-1]
if isBold:
tmp3 = []
tmp3html = []
for k in tmp2:
tmp3.append(r'\textbf{%s}' % k)
tmp3html.append('<b>%s</b>' % k)
tmp2 = tmp3
tmp2html = tmp3html
else:
tmp2html = []
tmp2html.extend(tmp2)
if not numpy.isnan(dispersion):
tmpdisp = dispersion/refalgert[j]
if tmpdisp >= maxfloatrepr or tmpdisp < 0.01:
tmpdisp = writeFEvalsMaxPrec(tmpdisp, precdispersion, maxfloatrepr=tmpdisp)
else:
tmpdisp = writeFEvalsMaxPrec(tmpdisp, precdispersion, maxfloatrepr=maxfloatrepr)
tmp2[-1] += (r'\mbox{\tiny (%s)}' % (tmpdisp))
tmp2html[-1] += ' (%s)' % tmpdisp
tmp2[-1] += str_significance_subsup
tmp2html[-1] += str_significance_subsup_html
curline.extend(tmp2)
tmp2html = ("").join(str(item) for item in tmp2html)
tmp2html = tmp2html.replace('$\infty$', '∞')
curlineHtml.append('<td sorttable_customkey=\"%f\">%s</td>' % (data, tmp2html))
curline.append('%d' % algnbsucc[i])
curline.append('/%d' % algnbruns[i])
table.append(curline)
curlineHtml.append('<td sorttable_customkey=\"%d\">%d/%d</td>\n' % (algnbsucc[i], algnbsucc[i], algnbruns[i]))
tableHtml.extend(curlineHtml[:])
extraeol.append('')
# Write table
res = tableXLaTeX(table, spec=spec, extraeol=extraeol)
try:
filename = os.path.join(outputdir, 'pptables_f%03d_%02dD.tex' % (df[1], df[0]))
f = open(filename, 'w')
f.write(header + '\n')
f.write(res)
res = ("").join(str(item) for item in tableHtml)
res = '\n<table class=\"sortable\" style=\"width:800px \">\n%s</table>\n<p/>\n' % res
if df[0] in (5, 20):
filename = os.path.join(outputdir, genericsettings.many_algorithm_file_name + '.html')
lines = []
with open(filename) as infile:
for line in infile:
if '<!--' + 'pptablesf%03d%02dDHtml' % (df[1], df[0]) + '-->' in line:
lines.append(res)
lines.append(line)
with open(filename, 'w') as outfile:
for line in lines:
outfile.write(line)
if verbose:
print 'Wrote table in %s' % filename
except:
raise
else:
f.close()
def main(dsList0, dsList1, minfvalue=1e-8, outputdir='', verbose=True):
"""Returns ERT1/ERT0 comparison figure."""
#plt.rc("axes", labelsize=20, titlesize=24)
#plt.rc("xtick", labelsize=20)
#plt.rc("ytick", labelsize=20)
#plt.rc("font", size=20)
#plt.rc("legend", fontsize=20)
# minfvalue = pproc.TargetValues.cast(minfvalue)
funInfos = ppfigparam.read_fun_infos(dsList0.isBiobjective())
dictFun0 = dsList0.dictByFunc()
dictFun1 = dsList1.dictByFunc()
for func in set.intersection(set(dictFun0), set(dictFun1)):
dictDim0 = dictFun0[func].dictByDim()
dictDim1 = dictFun1[func].dictByDim()
filename = os.path.join(outputdir,'ppfig2_f%03d' % (func))
dims = sorted(set.intersection(set(dictDim0), set(dictDim1)))
handles = []
dataperdim = {}
fvalueswitch = {}
nbtests = 0
for i, dim in enumerate(dimensions):
try:
entry0 = dictDim0[dim][0]
entry1 = dictDim1[dim][0]
except KeyError:
continue
nbtests += 1
# generateData:
data = _generateData(entry0, entry1, fthresh=fthresh)
dataperdim[dim] = data
if len(data[0]) == 0 and len(data[1]) == 0:
continue
# TODO: hack, modify slightly so line goes to 'zero'
if minfvalue:
for d in data:
tmp = d[:, 0]
tmp[tmp == 0] = min(min(tmp[tmp > 0]), minfvalue)**2
# plot
idx = np.isfinite(data[0][:, 1]) * np.isfinite(data[1][:, 1])
ydata = data[1][idx, 1]/data[0][idx, 1]
kwargs = styles[i].copy()
kwargs['label'] = '%2d-D' % dim
tmp = plotUnifLogXMarkers(data[0][idx, 0], ydata, nbperdecade=1, logscale=True, **kwargs)
plt.setp(tmp, markersize=3*linewidth)
plt.setp(tmp[0], ls='--')
# This is only one possibility:
#idx = (data[0][:, 3] >= 5) * (data[1][:, 3] >= 5)
idx = ((data[0][:, 1] <= 3 * np.median(entry0.maxevals))
* (data[1][:, 1] <= 3 * np.median(entry1.maxevals)))
if not idx.any():
fvalueswitch[dim] = np.inf
# Hack: fvalueswitch is the smallest value of f where the line
# was still solid.
continue
fvalueswitch[dim] = min(data[0][idx, 0])
ydata = data[1][idx, 1]/data[0][idx, 1]
tmp = plotUnifLogXMarkers(data[0][idx, 0], ydata, nbperdecade=1, logscale=True, **styles[i])
plt.setp(tmp[1], markersize=3*linewidth)
beautify(xmin=minfvalue)
#beautify()
ax = plt.gca()
# Freeze the boundaries
ax.set_autoscale_on(False)
#trans = transforms.blended_transform_factory(ax.transData, ax.transAxes)
# Plot everything else
for i, dim in enumerate(dimensions):
try:
entry0 = dictDim0[dim][0]
entry1 = dictDim1[dim][0]
data = dataperdim[dim]
except KeyError:
continue
if len(data[0]) == 0 and len(data[1]) == 0:
continue
# annotation
annotate(entry0, entry1, dim, minfvalue, nbtests=nbtests)
tmp0 = np.isfinite(data[0][:, 1])
tmp1 = np.isfinite(data[1][:, 1])
idx = tmp0 * tmp1
if not idx.any():
continue
#Do not plot anything else if it happens after minfvalue
if data[0][idx, 0][-1] <= minfvalue:
# hack for the legend
continue
# Determine which algorithm went further
algstoppedlast = 0
algstoppedfirst = 1
if np.sum(tmp0) < np.sum(tmp1):
algstoppedlast = 1
algstoppedfirst = 0
#marker if an algorithm stopped
ydata = data[1][idx, 1]/data[0][idx, 1]
plt.plot((data[0][idx, 0][-1], ), (ydata[-1], ), marker='D', ls='',
color=styles[i]['color'], markeredgecolor=styles[i]['color'],
markerfacecolor=styles[i]['color'], markersize=4*linewidth)
tmpy = ydata[-1]
# plot probability of success line
dataofinterest = data[algstoppedlast]
tmp = np.nonzero(idx)[0][-1] # Why [0]?
# add the last line for which both algorithm still have a success
idx = (data[algstoppedfirst][:, 2] == 0.) * (dataofinterest[:, 2] > 0.)
idx[tmp] = True
if np.sum(idx) <= 1:#len(idx) == 0 or not idx.any():
continue
ymin, ymax = plt.ylim()
#orientation = -1
ybnd = ymin
if algstoppedlast == 0:
ybnd = ymax
#orientation = 1
#ydata = orientation * dataofinterest[idx, 2] / 2 + 0.5
ydata = np.power(10, np.log10(ybnd) * (dataofinterest[idx, 2]
-offset*(5-i)*np.log10(ymax/ymin)/np.abs(np.log10(ybnd))))
ls = '-'
if dataofinterest[idx, 0][0] < fvalueswitch[dim]:
ls = '--'
tmp = plt.plot([dataofinterest[idx, 0][0]]*2, (tmpy, ydata[0]),
**styles[i])
plt.setp(tmp, ls=ls, marker='')
tmp = plt.plot((dataofinterest[idx, 0][0], ), (ydata[0], ), marker='D', ls='',
color=styles[i]['color'], markeredgecolor=styles[i]['color'],
markerfacecolor=styles[i]['color'], markersize=4*linewidth)
kwargs = styles[i].copy()
kwargs['ls'] = ls
tmp = plotUnifLogXMarkers(dataofinterest[idx, 0], ydata, nbperdecade=1, logscale=True, **kwargs)
plt.setp(tmp, markersize=3*linewidth)
#Do not plot anything else if it happens after minfvalue
if dataofinterest[idx, 0][-1] <= minfvalue:
continue
#plt.plot((dataofinterest[idx, 0][-1], ), (ydata[-1], ), marker='d',
# color=styles[i]['color'], markeredgecolor=styles[i]['color'],
# markerfacecolor=styles[i]['color'], markersize=4*linewidth)
if func in funInfos.keys():
plt.title(funInfos[func])
if func in functions_with_legend:
plt.legend(loc='best')
# save
saveFigure(filename, verbose=verbose)
plt.close()
def main(dsList0, dsList1, outputdir, verbose=True):
"""Generate a scatter plot figure.
TODO: """
#plt.rc("axes", labelsize=24, titlesize=24)
#plt.rc("xtick", labelsize=20)
#plt.rc("ytick", labelsize=20)
#plt.rc("font", size=20)
#plt.rc("legend", fontsize=20)
dictFunc0 = dsList0.dictByFunc()
dictFunc1 = dsList1.dictByFunc()
funcs = set(dictFunc0.keys()) & set(dictFunc1.keys())
if isinstance(targets, pproc.RunlengthBasedTargetValues):
linewidth = linewidth_rld_based
else:
linewidth = linewidth_default
funInfos = ppfigparam.read_fun_infos(dsList0.isBiobjective())
for f in funcs:
dictDim0 = dictFunc0[f].dictByDim()
dictDim1 = dictFunc1[f].dictByDim()
dims = set(dictDim0.keys()) & set(dictDim1.keys())
#set_trace()
for i, d in enumerate(dimensions):
try:
entry0 = dictDim0[d][0] # should be only one element
entry1 = dictDim1[d][0] # should be only one element
except (IndexError, KeyError):
continue
if linewidth: # plot all reliable ERT values as a line
all_targets = np.array(sorted(set(entry0.target).union(entry1.target), reverse=True))
assert entry0.detSuccessRates([all_targets[0]]) == 1.0
assert entry1.detSuccessRates([all_targets[0]]) == 1.0
all_targets = all_targets[np.where(all_targets <= targets((f, d))[0])[0]] #
xdata_all = np.array(entry0.detERT(all_targets))
ydata_all = np.array(entry1.detERT(all_targets))
# idx of reliable targets: last index where success rate >= 1/2 and ERT <= maxevals
idx = []
for ari in (np.where(entry0.detSuccessRates(all_targets) >= 0.5)[0],
np.where(entry1.detSuccessRates(all_targets) >= 0.5)[0],
np.where(xdata_all <= max(entry0.maxevals))[0],
np.where(ydata_all <= max(entry1.maxevals))[0]
):
if len(ari):
idx.append(ari[-1])
if len(idx) == 4:
max_idx = min(idx)
## at least up to the most difficult given target
## idx = max((idx, np.where(all_targets >= targets((f, d))[-1])[0][-1]))
xdata_all = xdata_all[:max_idx + 1]
ydata_all = ydata_all[:max_idx + 1]
idx = (numpy.isfinite(xdata_all)) * (numpy.isfinite(ydata_all))
assert idx.all()
if idx.any():
plt.plot(xdata_all[idx], ydata_all[idx], colors[i], ls='solid', lw=linewidth,
# TODO: ls has changed, check whether this works out
clip_on=False)
xdata = numpy.array(entry0.detERT(targets((f, d))))
ydata = numpy.array(entry1.detERT(targets((f, d))))
# plot "valid" data, those within maxevals
idx = np.logical_and(xdata < entry0.mMaxEvals(),
ydata < entry1.mMaxEvals())
# was:
# (numpy.isinf(xdata) == False) *
# (numpy.isinf(ydata) == False) *
# (xdata < entry0.mMaxEvals()) *
# (ydata < entry1.mMaxEvals()))
if idx.any():
try:
plt.plot(xdata[idx], ydata[idx], ls='',
markersize=markersize,
marker=markers[i], markerfacecolor='None',
markeredgecolor=colors[i], markeredgewidth=3,
clip_on=False)
except KeyError:
plt.plot(xdata[idx], ydata[idx], ls='', markersize=markersize,
marker='x', markerfacecolor='None',
markeredgecolor=colors[i], markeredgewidth=3,
clip_on=False)
#try:
# plt.scatter(xdata[idx], ydata[idx], s=10, marker=markers[i],
# facecolor='None', edgecolor=colors[i], linewidth=3)
#except ValueError:
# set_trace()
# plot beyond maxevals but finite data
idx = ((numpy.isinf(xdata) == False) *
(numpy.isinf(ydata) == False) *
np.logical_or(xdata >= entry0.mMaxEvals(),
ydata >= entry1.mMaxEvals()))
if idx.any():
try:
plt.plot(xdata[idx], ydata[idx], ls='',
markersize=markersize + markersize_addon_beyond_maxevals,
marker=markers[i], markerfacecolor='None',
markeredgecolor=colors[i], markeredgewidth=1,
clip_on=False)
except KeyError:
plt.plot(xdata[idx], ydata[idx], ls='', markersize=markersize,
marker='x', markerfacecolor='None',
markeredgecolor=colors[i], markeredgewidth=2,
clip_on=False)
#ax = plt.gca()
ax = plt.axes()
# plot data on the right edge
idx = numpy.isinf(xdata) * (numpy.isinf(ydata) == False)
if idx.any():
# This (seems to) transform inf to the figure limits!?
trans = blend(ax.transAxes, ax.transData)
#plt.scatter([1.]*numpy.sum(idx), ydata[idx], s=10, marker=markers[i],
# facecolor='None', edgecolor=colors[i], linewidth=3,
# transform=trans)
try:
plt.plot([1.]*numpy.sum(idx), ydata[idx],
markersize=markersize + markersize_addon_beyond_maxevals, ls='',
marker=markers[i], markerfacecolor='None',
markeredgecolor=colors[i], markeredgewidth=1,
transform=trans, clip_on=False)
except KeyError:
plt.plot([1.]*numpy.sum(idx), ydata[idx],
markersize=markersize, ls='',
marker='x', markerfacecolor='None',
markeredgecolor=colors[i], markeredgewidth=2,
transform=trans, clip_on=False)
#set_trace()
# plot data on the left edge
idx = (numpy.isinf(xdata)==False) * numpy.isinf(ydata)
if idx.any():
# This (seems to) transform inf to the figure limits!?
trans = blend(ax.transData, ax.transAxes)
# plt.scatter(xdata[idx], [1.-offset]*numpy.sum(idx), s=10, marker=markers[i],
# facecolor='None', edgecolor=colors[i], linewidth=3,
# transform=trans)
try:
plt.plot(xdata[idx], [1.-offset]*numpy.sum(idx),
markersize=markersize + markersize_addon_beyond_maxevals, ls='',
marker=markers[i], markerfacecolor='None',
markeredgecolor=colors[i], markeredgewidth=1,
transform=trans, clip_on=False)
except KeyError:
plt.plot(xdata[idx], [1.-offset]*numpy.sum(idx),
markersize=markersize, ls='',
marker='x', markerfacecolor='None',
markeredgecolor=colors[i], markeredgewidth=2,
transform=trans, clip_on=False)
# plot data in the top corner
idx = numpy.isinf(xdata) * numpy.isinf(ydata)
if idx.any():
# plt.scatter(xdata[idx], [1.-offset]*numpy.sum(idx), s=10, marker=markers[i],
# facecolor='None', edgecolor=colors[i], linewidth=3,
# transform=trans)
try:
plt.plot([1.-offset]*numpy.sum(idx), [1.-offset]*numpy.sum(idx),
markersize=markersize + markersize_addon_beyond_maxevals, ls='',
marker=markers[i], markerfacecolor='None',
markeredgecolor=colors[i], markeredgewidth=1,
transform=ax.transAxes, clip_on=False)
except KeyError:
plt.plot([1.-offset]*numpy.sum(idx), [1.-offset]*numpy.sum(idx),
markersize=markersize, ls='',
marker='x', markerfacecolor='None',
markeredgecolor=colors[i], markeredgewidth=2,
transform=ax.transAxes, clip_on=False)
#set_trace()
beautify()
for i, d in enumerate(dimensions):
try:
entry0 = dictDim0[d][0] # should be only one element
entry1 = dictDim1[d][0] # should be only one element
except (IndexError, KeyError):
continue
minbnd, maxbnd = plt.xlim()
plt.plot((entry0.mMaxEvals(), entry0.mMaxEvals()),
# (minbnd, entry1.mMaxEvals()), ls='-', color=colors[i],
(max([minbnd, entry1.mMaxEvals()/max_evals_line_length]), entry1.mMaxEvals()), ls='-', color=colors[i],
zorder=-1)
plt.plot(# (minbnd, entry0.mMaxEvals()),
(max([minbnd, entry0.mMaxEvals()/max_evals_line_length]), entry0.mMaxEvals()),
(entry1.mMaxEvals(), entry1.mMaxEvals()), ls='-',
color=colors[i], zorder=-1)
plt.xlim(minbnd, maxbnd)
plt.ylim(minbnd, maxbnd)
#Set the boundaries again: they changed due to new plots.
#plt.axvline(entry0.mMaxEvals(), ls='--', color=colors[i])
#plt.axhline(entry1.mMaxEvals(), ls='--', color=colors[i])
if f in funInfos.keys():
plt.ylabel(funInfos[f])
filename = os.path.join(outputdir, 'ppscatter_f%03d' % f)
saveFigure(filename, verbose=verbose)
if f == 1:
algName1 = toolsdivers.str_to_latex(toolsdivers.strip_pathname1(entry1.algId))
algName0 = toolsdivers.str_to_latex(toolsdivers.strip_pathname1(entry0.algId))
save_single_functions_html(
os.path.join(outputdir, genericsettings.two_algorithm_file_name),
"%s vs %s" % (algName1, algName0),
algorithmCount=AlgorithmCount.TWO)
plt.close()