def plotmntesterrperfile(jsonfile,testfile, desc,folder):
minp = np.inf
minq = np.inf
maxp = 0
maxq = 0
miny0 = 0
minl1 = np.inf
minl2 = np.inf
minlinf = np.inf
pp = 0
qq =0
outfile1 = folder+"/"+desc+".299445.png"
# outfile2 = folder+"/"+fno+"_index.299445.png"
X_test, Y_test = readData(testfile)
import json
if jsonfile:
with open(jsonfile, 'r') as fn:
datastore = json.load(fn)
from apprentice import monomial
import apprentice
for key in sorted(datastore.keys()):
pdeg = datastore[key]['m']
qdeg = datastore[key]['n']
if(pdeg<minp):
minp=pdeg
if(qdeg<minq):
minq=qdeg
if pdeg > maxp:
maxp = pdeg
if qdeg > maxq:
maxq = qdeg
# print(minp,maxp,minq,maxq)
error = np.zeros(shape = (maxp-minp+1,maxq-minq+1))
for key in sorted(datastore.keys()):
pdeg = datastore[key]['m']
qdeg = datastore[key]['n']
Y_pred = np.array([],dtype=np.float64)
if('scaler' in datastore[key]):
rappsip = RationalApproximationSIP(datastore[key])
Y_pred = rappsip.predictOverArray(X_test)
# print(Y_pred)
else:
structp = apprentice.monomialStructure(datastore[key]['dim'], pdeg)
structq = apprentice.monomialStructure(datastore[key]['dim'], qdeg)
for x in X_test:
nnn = np.array(monomial.recurrence(x, structp))
p = np.array(datastore[key]['pcoeff']).dot(nnn)
ddd = np.array(monomial.recurrence(x, structq))
q = np.array(datastore[key]['qcoeff']).dot(ddd)
Y_pred = np.append(Y_pred,(p/q))
# print(np.c_[Y_pred,Y_test])
l1 = np.sum(np.absolute(Y_pred-Y_test))
# print(l1)
l2 = np.sqrt(np.sum((Y_pred-Y_test)**2))
linf = np.max(np.absolute(Y_pred-Y_test))
x000 = np.zeros(datastore[key]['dim'])
y000 = rappsip.predict(x000)
print("p%d q%d %f"%(pdeg,qdeg,y000))
error[pdeg-minp][qdeg-minq] = l2
if(minl2>l2):
minl2 = l2
minl1 = l1
minlinf = linf
print(linf)
pp = pdeg
qq = qdeg
miny0 = y000
# print(miiny0)
import matplotlib as mpl
import matplotlib.pyplot as plt
mpl.rc('text', usetex = True)
mpl.rc('font', family = 'serif', size=12)
mpl.style.use("ggplot")
cmapname = 'viridis'
plt.clf()
print(error)
markersize = 1000
vmin = -4
vmax = 2
X,Y = np.meshgrid(range(minq,maxq+1),range(minp,maxp+1))
# plt.scatter(X,Y , marker = 's', s=markersize, c = np.ma.log10(error), cmap = cmapname, vmin=vmin, vmax=vmax, alpha = 1)
plt.scatter(X,Y , marker = 's', s=markersize, c = error, cmap = cmapname, alpha = 1)
plt.xlabel("$n$")
plt.ylabel("$m$")
plt.xlim((minq-1,maxq+1))
plt.ylim((minp-1,maxp+1))
b=plt.colorbar()
# b.set_label("$\log_{10}\\left|\\left|f - \\frac{p^m}{q^n}\\right|\\right|_2$")
b.set_label("$\\left|\\left|f - \\frac{p^m}{q^n}\\right|\\right|_2$")
plt.title("l1=%f, l2=%f, linf=%f y0=%f found at (%d,%d)"%(minl1,minl2,minlinf,pp,qq,miny0))
plt.savefig(outfile1)
def plot2Dsurface(fname, noise, m, n, ts, fndesc, papp_or_no_papp):
noisestr = ""
if (noise != "0"):
noisestr = "_noisepct" + noise
folder = "%s%s_%s" % (fname, noisestr, ts)
if not os.path.exists(folder + "/plots"):
os.mkdir(folder + '/plots')
if (papp_or_no_papp == "papp"):
cols = 4
elif (papp_or_no_papp == "no_papp"):
cols = 3
else:
raise Exception("papp_or_no_papp ambiguous")
box = getbox(fname)
if (len(box) != 2):
print("{} cannot handle dim != 2. Box len was {}".format(
sys.argv[0], len(box)))
sys.exit(1)
npoints = 250
X_test1 = np.linspace(box[0][0], box[0][1], num=npoints)
X_test2 = np.linspace(box[1][0], box[1][1], num=npoints)
outx1 = "%s/plots/Cfnsurf_X_%s%s_p%s_q%s_ts%s.csv" % (folder, fname,
noisestr, m, n, ts)
outy = "%s/plots/Cfnsurf_Y_%s%s_p%s_q%s_ts%s.csv" % (folder, fname,
noisestr, m, n, ts)
rappsipfile = "%s/out/%s%s_%s_p%s_q%s_ts%s.json" % (folder, fname,
noisestr, ts, m, n, ts)
rappfile = "%s/outra/%s%s_%s_p%s_q%s_ts%s.json" % (folder, fname, noisestr,
ts, m, n, ts)
pappfile = "%s/outpa/%s%s_%s_p%s_q%s_ts%s.json" % (folder, fname, noisestr,
ts, m, n, ts)
if not os.path.exists(rappsipfile):
print("rappsipfile %s not found" % (rappsipfile))
exit(1)
if not os.path.exists(rappfile):
print("rappfile %s not found" % (rappfile))
exit(1)
if not os.path.exists(pappfile):
print("pappfile %s not found" % (pappfile))
exit(1)
rappsip = RationalApproximationSIP(rappsipfile)
rapp = RationalApproximationONB(fname=rappfile)
papp = PolynomialApproximation(fname=pappfile)
if (rappsip.dim != 2):
print("{} cannot handle dim != 2 Dim found in datastore was {}".format(
sys.argv[0], rappsip.dim))
sys.exit(1)
from apprentice import testData
Y_test = []
for x in X_test1:
for y in X_test2:
Y_test.append(
eval('testData.' + fname + '([' + str(x) + ',' + str(y) +
'])'))
Y_pred_rappsip = []
for x in X_test1:
for y in X_test2:
Y_pred_rappsip.append(rappsip.predict([x, y]))
Y_pred_rapp = []
for x in X_test1:
for y in X_test2:
Y_pred_rapp.append(rapp([x, y]))
Y_pred_papp = []
if (papp_or_no_papp == "papp"):
for x in X_test1:
for y in X_test2:
Y_pred_papp.append(papp([x, y]))
np.savetxt(outx1, np.stack((X_test1, X_test2), axis=1), delimiter=",")
def removezeros(YYY):
for num, y in enumerate(YYY):
if (y == 0):
# print(y, num)
lower = 0
upper = len(YYY) - 1
index = num
index -= 1
while (index >= 0):
if (YYY[index] != 0):
lower = index
break
index -= 1
index = num
index += 1
while (index < len(YYY)):
if (YYY[index] != 0):
upper = index
break
index += 1
YYY[num] = (YYY[lower] + YYY[upper]) / 2
# print("became")
# print(YYY[num], num)
return YYY
Y_pred_papp_diff = np.absolute(np.array(Y_pred_papp) - np.array(Y_test))
Y_pred_papp_diff = removezeros(Y_pred_papp_diff)
Y_pred_rapp_diff = np.absolute(np.array(Y_pred_rapp) - np.array(Y_test))
Y_pred_rapp_diff = removezeros(Y_pred_rapp_diff)
Y_pred_rappsip_diff = np.absolute(
np.array(Y_pred_rappsip) - np.array(Y_test))
Y_pred_rappsip_diff = removezeros(Y_pred_rappsip_diff)
np.savetxt(
outy,
np.stack((np.ma.log10(Y_pred_papp_diff), np.ma.log10(Y_pred_rapp_diff),
np.ma.log10(Y_pred_rappsip_diff)),
axis=1),
delimiter=",")
minyval = np.inf
maxyval = 0
arr1 = np.array([min(Y_pred_rappsip), min(Y_test), min(Y_pred_rapp)])
if (papp_or_no_papp == "papp"):
np.append(arr1, min(Y_pred_papp))
minyval = np.min(arr1)
arr1 = np.array([max(Y_pred_rappsip), max(Y_test), max(Y_pred_rapp)])
if (papp_or_no_papp == "papp"):
np.append(arr1, max(Y_pred_papp))
maxyval = np.max(arr1)
import matplotlib
import matplotlib.colors as colors
import matplotlib as mpl
import matplotlib.pyplot as plt
import matplotlib.text as text
cmap1 = matplotlib.cm.RdYlBu
cmap3 = matplotlib.cm.seismic
cmap4 = matplotlib.cm.viridis
cmap5 = matplotlib.cm.coolwarm
cmap6 = matplotlib.cm.magma
cmap2 = 'hot'
# https://stackoverflow.com/questions/16267143/matplotlib-single-colored-colormap-with-saturation
from matplotlib.colors import LinearSegmentedColormap
def CustomCmap(from_rgb, to_rgb):
# from color r,g,b
r1, g1, b1 = from_rgb
# to color r,g,b
r2, g2, b2 = to_rgb
cdict = {
'red': ((0, r1, r1), (1, r2, r2)),
'green': ((0, g1, g1), (1, g2, g2)),
'blue': ((0, b1, b1), (1, b2, b2))
}
ccc = LinearSegmentedColormap('custom_cmap', cdict)
return ccc
cmap7 = CustomCmap([0.00, 0.00, 0.00], [0.02, 0.75, 1.00])
cmap8 = CustomCmap([1.00, 0.42, 0.04], [0.02, 0.75, 1.00])
cmap = cmap8
fig = plt.figure(figsize=(20, 5))
ax = fig.add_subplot(1, cols, 1, projection='3d')
Y_test = np.reshape(Y_test, [len(X_test1), len(X_test2)])
im1 = ax.contour3D(X_test1,
X_test2,
Y_test,
100,
cmap=cmap,
norm=colors.SymLogNorm(linthresh=0.2,
linscale=0.5,
vmin=minyval,
vmax=maxyval),
alpha=0.8)
ax.set_title('(a)', fontsize=12)
ax.set_xlabel('$x_1$', fontsize=12)
ax.set_ylabel('$x_2$', fontsize=12)
ax = fig.add_subplot(1, cols, 2, projection='3d')
Y_pred_rappsip = np.reshape(Y_pred_rappsip, [len(X_test1), len(X_test2)])
im1 = ax.contour3D(X_test1,
X_test2,
Y_pred_rappsip,
100,
cmap=cmap,
norm=colors.SymLogNorm(linthresh=0.2,
linscale=0.5,
vmin=minyval,
vmax=maxyval),
alpha=0.8)
ax.set_title('(b)', fontsize=12)
ax.set_xlabel('$x_1$', fontsize=12)
ax.set_ylabel('$x_2$', fontsize=12)
ax = fig.add_subplot(1, cols, 3, projection='3d')
Y_pred_rapp = np.reshape(Y_pred_rapp, [len(X_test1), len(X_test2)])
im1 = ax.contour3D(X_test1,
X_test2,
Y_pred_rapp,
100,
cmap=cmap,
norm=colors.SymLogNorm(linthresh=0.2,
linscale=0.5,
vmin=minyval,
vmax=maxyval),
alpha=0.8)
ax.set_title('(c)', fontsize=12)
ax.set_xlabel('$x_1$', fontsize=12)
ax.set_ylabel('$x_2$', fontsize=12)
if (papp_or_no_papp == "papp"):
ax = fig.add_subplot(1, cols, 4, projection='3d')
Y_pred_papp = np.reshape(Y_pred_papp, [len(X_test1), len(X_test2)])
im1 = ax.contour3D(X_test1,
X_test2,
Y_pred_papp,
100,
cmap=cmap,
norm=colors.SymLogNorm(linthresh=0.2,
linscale=0.5,
vmin=minyval,
vmax=maxyval),
alpha=0.8)
ax.set_title('(d)', fontsize=12)
ax.set_xlabel('$x_1$', fontsize=12)
ax.set_ylabel('$x_2$', fontsize=12)
mmm = plt.cm.ScalarMappable(cmap=cmap)
mmm.set_array(Y_pred_rapp)
mmm.set_clim(minyval, maxyval)
b2 = fig.colorbar(mmm)
b2.set_label("$\\frac{p(x_1,x_2)}{q(x_1,x_2)}$", fontsize=16)
outfile = "%s/plots/P2d1f_%s%s_p%s_q%s_ts%s_%s.pdf" % (
folder, fname, noisestr, m, n, ts, papp_or_no_papp)
# plt.show()
plt.savefig(outfile)
print("open %s;" % (outfile))
plt.close('all')
def plotoptiterationmaps(farr, noisearr, ts):
import glob
import json
import re
import json
openfileStr = ""
lx = "$\\log_{10}(\\alpha(M) + \\alpha(N))$"
ly = "$\\log_{10}(\\Delta_{MN})$"
logy = True
logx = True
for num, fname in enumerate(farr):
box = getbox(fname)
if (len(box) != 2):
print("{} cannot handle dim != 2. Box len was {}".format(
sys.argv[0], len(box)))
sys.exit(1)
npoints = 1000
X_test1 = np.linspace(box[0][0], box[0][1], num=npoints)
X_test2 = np.linspace(box[1][0], box[1][1], num=npoints)
for noise in noisearr:
noisestr = ""
if (noise != "0"):
noisestr = "_noisepct" + noise
folder = "%s%s_%s" % (fname, noisestr, ts)
if not os.path.exists(folder):
print("Folder '{}' not found.".format(folder))
sys.exit(1)
if not os.path.exists(folder + "/plots"):
os.mkdir(folder + '/plots')
filelist = np.array(glob.glob(folder + "/out/*.json"))
filelist = np.sort(filelist)
for file in filelist:
if file:
with open(file, 'r') as fn:
datastore = json.load(fn)
if (datastore["dim"] != 2):
print(
"{} cannot handle dim != 2 Dim found in datastore was {}"
.format(sys.argv[0], datastore["dim"]))
sys.exit(1)
if (len(datastore["iterationinfo"]) != 3):
continue
# if(datastore["m"] !=5 or datastore["n"] !=3):
# continue
print(file)
maxy_pq = 0
miny_pq = np.inf
maxy_q = 0
miny_q = np.inf
data = {}
m = datastore["m"]
n = datastore["n"]
outx1 = "%s/plots/Cimap_X_%s%s_p%d_q%d_ts%s.csv" % (
folder, fname, noisestr, m, n, ts)
outy_pq = "%s/plots/Cimap_Y_pq_%s%s_p%d_q%d_ts%s.csv" % (
folder, fname, noisestr, m, n, ts)
outy_q = "%s/plots/Cimap_Y_q_%s%s_p%d_q%d_ts%s.csv" % (
folder, fname, noisestr, m, n, ts)
polex = "%s/plots/Cimap_pole_x_%s%s_p%d_q%d_ts%s.csv" % (
folder, fname, noisestr, m, n, ts)
for iterno in range(3):
data[iterno] = {}
if file:
with open(file, 'r') as fn:
datastore = json.load(fn)
ii = datastore["iterationinfo"]
datastore['pcoeff'] = ii[iterno]['pcoeff']
datastore['qcoeff'] = ii[iterno]['qcoeff']
iro = ii[iterno]['robOptInfo']['robustArg']
iro.append(ii[iterno]['robOptInfo']['robustObj'])
data[iterno]['robarg'] = iro
rappsip = RationalApproximationSIP(datastore)
Y_pred_pq = []
for x in X_test1:
for y in X_test2:
Y_pred_pq.append(rappsip.predict([x, y]))
mmm = max(Y_pred_pq)
# print(mmm)
if (mmm > maxy_pq):
maxy_pq = mmm
mmm = min(Y_pred_pq)
# print(mmm)
if (mmm < miny_pq):
miny_pq = mmm
data[iterno]['Y_pred_pq'] = Y_pred_pq
Y_pred_q = []
for x in X_test1:
for y in X_test2:
X = [x, y]
X = rappsip._scaler.scale(np.array(X))
Y_pred_q.append(rappsip.denom(X))
mmm = max(Y_pred_q)
# print(mmm)
if (mmm > maxy_q):
maxy_q = mmm
mmm = min(Y_pred_q)
# print(mmm)
if (mmm < miny_q):
miny_q = mmm
data[iterno]['Y_pred_q'] = Y_pred_q
np.savetxt(outx1,
np.stack((X_test1, X_test2), axis=1),
delimiter=",")
np.savetxt(outy_pq,
np.stack(
(data[0]['Y_pred_pq'], data[1]['Y_pred_pq'],
data[2]['Y_pred_pq']),
axis=1),
delimiter=",")
np.savetxt(outy_q,
np.stack((data[0]['Y_pred_q'], data[1]['Y_pred_q'],
data[2]['Y_pred_q']),
axis=1),
delimiter=",")
np.savetxt(polex,
np.stack((data[0]['robarg'], data[1]['robarg'],
data[2]['robarg']),
axis=1),
delimiter=",")
import matplotlib
import matplotlib.colors as colors
import matplotlib as mpl
import matplotlib.pyplot as plt
import matplotlib.text as text
cmap1 = matplotlib.cm.RdYlBu
cmap3 = matplotlib.cm.seismic
cmap4 = matplotlib.cm.viridis
cmap5 = matplotlib.cm.coolwarm
cmap6 = matplotlib.cm.magma
cmap2 = 'hot'
x1lim = (box[0][0], box[0][1])
x2lim = (box[1][0], box[1][1])
fig = plt.figure(figsize=(17, 8))
ypredmaster = []
from matplotlib.colors import LinearSegmentedColormap
def CustomCmap(from_rgb, to_rgb):
# from color r,g,b
r1, g1, b1 = from_rgb
# to color r,g,b
r2, g2, b2 = to_rgb
cdict = {
'red': ((0, r1, r1), (1, r2, r2)),
'green': ((0, g1, g1), (1, g2, g2)),
'blue': ((0, b1, b1), (1, b2, b2))
}
ccc = LinearSegmentedColormap('custom_cmap', cdict)
return ccc
cmap7 = CustomCmap([0.00, 0.00, 0.00], [0.02, 0.75, 1.00])
cmap8 = CustomCmap([1.00, 0.42, 0.04], [0.02, 0.75, 1.00])
cmap = cmap8
# axarray = np.array([])
for iterno in range(3):
ax = fig.add_subplot(2, 3, iterno + 1, projection='3d')
min111 = min(data[iterno]['Y_pred_pq'])
max111 = max(data[iterno]['Y_pred_pq'])
Y_pred_pq = np.reshape(
data[iterno]['Y_pred_pq'],
[len(X_test1), len(X_test2)])
# im1 = ax.contour3D(X_test1, X_test2, Y_pred_pq, 100, cmap=cmap2, norm = colors.SymLogNorm(linthresh=1, linscale=1,vmin=miny_pq, vmax=maxy_pq),alpha=0.8)
im1 = ax.contour3D(X_test1,
X_test2,
Y_pred_pq,
100,
cmap=cmap,
norm=colors.SymLogNorm(linthresh=0.2,
linscale=0.5,
vmin=miny_pq,
vmax=maxy_pq),
alpha=0.8)
# im1 = ax.contour3D(X_test1, X_test2, Y_pred_pq, cmap=matplotlib.cm.seismic,vmin=miny_pq, vmax=maxy_pq)
# axarray = np.append(axarray,ax)
if (iterno == 0):
ypredmaster = Y_pred_pq
if (iterno == 2):
mmm = plt.cm.ScalarMappable(cmap=cmap)
mmm.set_array(ypredmaster)
mmm.set_clim(miny_pq, maxy_pq)
b1 = fig.colorbar(mmm)
b1.set_label("$\\frac{p(x_1,x_2)}{q(x_1,x_2)}$",
fontsize=16)
# ax.set_title('Iteration: %d'%(iterno+1), fontsize = 12)
ax.set_xlabel('$x_1$', fontsize=14)
ax.set_ylabel('$x_2$', fontsize=14)
# b1 = fig.colorbar(im1,ax=ax, shrink=0.95,extend='both')
# if(iterno ==2):
# b1.set_label("$\\frac{p_m(x_1,x_2)}{q_n(x_1,x_2)}$", fontsize = 14)
ax.set(xlim=x1lim, ylim=x2lim)
ax.view_init(60, 35)
if (iterno != 2):
ax.scatter(data[iterno]['robarg'][1],
data[iterno]['robarg'][0],
marker='*',
c="black",
s=333,
alpha=1,
zorder=1)
for iterno in range(3):
ax = fig.add_subplot(2, 3, iterno + 4, projection='3d')
min111 = min(data[iterno]['Y_pred_pq'])
max111 = max(data[iterno]['Y_pred_pq'])
Y_pred_q = np.reshape(
data[iterno]['Y_pred_q'],
[len(X_test1), len(X_test2)])
im2 = ax.contour3D(X_test1,
X_test2,
Y_pred_q,
cmap=cmap,
norm=colors.SymLogNorm(linthresh=4,
linscale=0.6,
vmin=miny_q,
vmax=maxy_q))
# im2 = ax.contour3D(X_test1, X_test2, Y_pred_q, cmap=cmap2,vmin=miny_q, vmax=maxy_q)
# ax.set_title('Iteration: %d'%(iterno+1), fontsize = 12)
ax.set_xlabel('$x_1$', fontsize=14)
ax.set_ylabel('$x_2$', fontsize=14)
# b2 = fig.colorbar(im2,ax=ax, shrink=0.95,extend='both')
if (iterno == 0):
ypredmaster = Y_pred_q
if (iterno == 2):
mmm = plt.cm.ScalarMappable(cmap=cmap)
mmm.set_array(ypredmaster)
mmm.set_clim(miny_q, maxy_q)
b2 = fig.colorbar(mmm)
b2.set_label("$q(x_1,x_2)$", fontsize=12)
ax.view_init(60, 35)
if (iterno != 2):
ax.scatter(data[iterno]['robarg'][1],
data[iterno]['robarg'][0],
marker='*',
c="black",
s=333,
alpha=1,
zorder=1)
# b2=fig.colorbar(im2,ax=ax, shrink=0.95,extend='both')
# fig.suptitle("%s%s. m = %d, n = %d. trainingsize = %s "%(fname, noisestr,m,n,ts), fontsize = 18)
# plt.show()
# exit(1)
#
# cmap2='hot'
# import matplotlib
# import matplotlib.colors as colors
# cmap1 = matplotlib.cm.RdYlBu
# import matplotlib as mpl
# import matplotlib.pyplot as plt
# import matplotlib.text as text
# fig, axarr = plt.subplots(2,3, figsize=(17,8))
# x1lim = (box[0][0]-0.5,box[0][1]+0.5)
# x2lim = (box[1][0]-0.5,box[1][1]+0.5)
# # if(abs(miny_pq) < abs(maxy_pq)):
# # miny_pq = -abs(maxy_pq)
# # maxy_pq = abs(maxy_pq)
# # elif(abs(miny_pq) > abs(maxy_pq)):
# # miny_pq = -abs(miny_pq)
# # maxy_pq = abs(miny_pq)
# # xx,yy = np.meshgrid(X_test1,X_test2)
# for iterno in range(3):
# min111 = min(data[iterno]['Y_pred_pq'])
# max111 = max(data[iterno]['Y_pred_pq'])
# # print(data[iterno]['Y_pred'])
# Y_pred_pq = np.reshape(data[iterno]['Y_pred_pq'], [len(X_test1), len(X_test2)])
# # print(Y_pred_pq)
# # im1 = axarr[0][iterno].pcolormesh(X_test1, X_test2, Y_pred_pq, cmap=cmap, vmin=min111, vmax=max111)
# # im1 = axarr[0][iterno].pcolormesh(X_test1,X_test2, Y_pred_pq, cmap=cmap)
# # im1 = axarr[0][iterno].pcolormesh(X_test1, X_test2, Y_pred_pq)
# # im1 = axarr[0][iterno].pcolormesh(X_test1, X_test2, Y_pred_pq, cmap=cmap1, vmin=miny_pq, vmax=maxy_pq)
# im1 = axarr[0][iterno].pcolormesh(X_test1, X_test2, Y_pred_pq, cmap=cmap1, norm = colors.SymLogNorm(linthresh=0.1, linscale=1,vmin=miny_pq, vmax=maxy_pq))
# # im1 = axarr[0][iterno].imshow(Y_pred_pq, extent=[-1, 1, -1, 1], cmap=cmap1)
# # axarr[0][iterno].axis(aspect='image')
# axarr[0][iterno].set_xlabel('$x_1$', fontsize = 14)
# axarr[0][iterno].set_ylabel('$x_2$', fontsize = 14)
# axarr[0][iterno].set(xlim=x1lim,ylim=x2lim)
# if(iterno !=2):
# axarr[0][iterno].scatter(data[iterno]['robarg'][1], data[iterno]['robarg'][0], marker = 'x', c = "black" ,s=50, alpha = 1)
# # if(iterno ==0):
# # fig.colorbar(im1)
# # axarr[0][iterno].axis('tight')
#
# Y_pred_q = np.reshape(data[iterno]['Y_pred_q'], [len(X_test1), len(X_test2)])
# # im2 = axarr[1][iterno].contour(X_test1, X_test2, Y_pred_q, cmap=cmap2, vmin=miny_q, vmax=maxy_q)
# im2 = axarr[1][iterno].pcolormesh(X_test1, X_test2, Y_pred_q, cmap=cmap1, norm = colors.SymLogNorm(linthresh=4, linscale=0.6,vmin=miny_q, vmax=maxy_q))
# axarr[1][iterno].set_xlabel('$x_1$', fontsize = 14)
# axarr[1][iterno].set_ylabel('$x_2$', fontsize = 14)
# axarr[1][iterno].set(xlim=x1lim,ylim=x2lim)
#
# if(iterno !=2):
# axarr[1][iterno].scatter(data[iterno]['robarg'][1], data[iterno]['robarg'][0], marker = 'x', c = "black" ,s=50, alpha = 1)
# for ax in axarr.flat:
# ax.label_outer()
# b1=fig.colorbar(im1,ax=axarr[0].ravel().tolist(), shrink=0.95,extend='both')
# b1.set_label("$\\frac{p_m(x_1,x_2)}{q_n(x_1,x_2)}$", fontsize = 16)
# b2=fig.colorbar(im2,ax=axarr[1].ravel().tolist(), shrink=0.95,extend='both')
# b2.set_label("$q_n(x_1,x_2)$", fontsize = 16)
#
# fig.suptitle("%s%s. m = %d, n = %d. trainingsize = %s "%(fname, noisestr,m,n,ts), fontsize = 18)
if not os.path.exists(folder + "/plots"):
os.mkdir(folder + '/plots')
outfile = "%s/plots/Pimap_%s%s_p%d_q%d_ts%s.pdf" % (
folder, fname, noisestr, m, n, ts)
plt.savefig(outfile)
plt.clf()
openfileStr += "open " + outfile + "; "
print(openfileStr)
plt.close('all')
print(openfileStr)