def get_R2(bass=None):
# pkl_file = open('standardn3')
pkl_file = open('PD6--RM---')
data = pickle.load(pkl_file)
msrt = data['MSRT']
# fn = "F:\MARS\data_save\STDAND/zhi10-5vs1.CDF"
fn = "F:\MARS\data_save\D6/73.CDF"
options = data['options']
ncr = netcdf_reader(fn, bmmap=False)
pw = options['pw']
w = options['w']
thre = options['thres']
rts = msrt['rt']
mss = msrt['ms']
mz = msrt['mz']
segno = msrt['segno']
results = data['results']
inds = [75]
tzs = results[0]['tzs'][inds, :]
RG = np.arange(np.min(tzs), np.max(tzs))
dd = np.zeros((len(RG), len(mz)))
xx = ncr.mat(int(RG[0]), int(RG[-1]), 1)
for j in inds:
newd = np.zeros((len(RG), len(mz)))
tz, seg, x, segx, maxpos, irr = findTZ(ncr, rts[j],
np.array(mss[j], ndmin=2), pw,
w, 'RM')
c, vsel, pc = gridsearch(x, maxpos, 6, irr)
chrom, area, heig, d = polyfitting(c, x, vsel, np.array(mss[j],
ndmin=2))
re = np.searchsorted(RG, [tz[0], tz[1]])
newd[re[0]:re[1], :] = d
dd = dd + newd
xx1 = ncr.mat(int(RG[0]) - 8, int(RG[-1]) + 8, 1)
seg = np.array([[0, 8], [xx1['d'].shape[0] - 8, xx1['d'].shape[0]]])
fit, bas = backremv(xx1['d'], seg)
bbas = bas[8:xx1['d'].shape[0] - 8, :]
if bass == 'baseline':
xxo = xx['d'] - bbas
resn = xx['d'] - bbas - dd
else:
xxo = xx['d']
resn = xx['d'] - dd
un = np.sum((np.power(resn, 2)))
sstn = np.sum(np.power(xxo, 2))
R2 = (sstn - un) / sstn
print(R2)
plt.plot(np.sum(xx1['d'], axis=1), 'b')
# plt.plot(range(8,xx1['d'].shape[0]-8),np.sum(dd,axis=0),'r')
t = range(8, xx1['d'].shape[0] - 8)
plt.plot(t, np.sum(dd, axis=1), 'r')
plt.plot(np.sum(bas, axis=1), 'k')
plt.show()
def plot_compara1():
pkl_file = open('PD6--RM--')
data = pickle.load(pkl_file)
msrt = data['MSRT']
fn = "F:\MARS\data_save\D6/75.CDF"
options = data['options']
ncr = netcdf_reader(fn, bmmap=False)
pw = options['pw']
w = options['w']
thre = options['thres']
rts = msrt['rt']
ms = msrt['ms']
mz = msrt['mz']
# com = [79]
com = range(22, 27)
rttic = ncr.tic()['rt']
plt.figure()
ax = plt.subplot(111)
rang = []
colors = ['r', 'g', 'c', 'y', 'm']
for ii, i in enumerate(com):
tz, seg2, x, segx, maxpos, irr = findTZ(ncr,
rts[i],
np.array(ms[i], ndmin=2),
pw,
w,
mth='RM')
rang.extend(tz)
c, vsel, pc = gridsearch(x, maxpos, 6, irr)
chrom, area, heig, d = polyfitting(c, x, vsel, np.array(ms[i],
ndmin=2))
# plt.subplot(111)
# plt.plot(np.sum(x, axis=1))
plt.plot(rttic[range(tz[0], tz[1])],
chrom,
color=colors[ii],
linewidth=3.0,
label='c' + str(ii + 1))
plt.plot(
rttic[range(tz[0], tz[1])],
chrom,
color='m',
)
segx = ncr.mat(min(rang), max(rang) - 1, 1)
rtseg = rttic[range(min(rang), max(rang))]
plt.plot(rtseg,
np.sum(segx['d'], axis=1),
color='b',
linewidth=3.0,
label='raw')
ax.set_xlim((min(rtseg), max(rtseg)))
ax.set_xlabel('retention time (min)')
ax.set_ylabel('intensity')
# plt.legend(loc='upper left')
plt.show()
def plot_RM():
pkl_file = open('PD6--RM--')
data = pickle.load(pkl_file)
msrt = data['MSRT']
fn = "F:\MARS\data_save\D6/73.CDF"
options = data['options']
ncr = netcdf_reader(fn, bmmap=False)
pw = options['pw']
w = options['w']
thre = options['thres']
rts = msrt['rt']
ms = msrt['ms']
mz = msrt['mz']
com = 33
rttic = ncr.tic()['rt']
tz, seg1, x, segx, maxpos, irr = findTZ(ncr,
rts[com],
np.array(ms[com], ndmin=2),
pw,
w,
mth='MSWFA')
resp1, eig1, ir = MSWFA(segx,
np.array(ms[com], ndmin=2),
5,
w,
mth='MSWFA')
tz, seg2, x, segx, maxpos, irr = findTZ(ncr,
rts[com],
np.array(ms[com], ndmin=2),
pw,
w,
mth='RM')
resp2, eig2, ir = MSWFA(segx, np.array(ms[com], ndmin=2), 5, w, mth='RM')
po = rttic[tz]
x1 = rttic[range(seg1[0], seg1[1])]
x2 = rttic[range(seg2[0], seg2[1])]
pos = np.searchsorted(x2, po)
plt.figure()
ax1 = plt.subplot(211)
ax1.plot(x1, eig1, color='r', label='MSWFA')
ax1.plot(x2, eig2, color='b', label='RM')
ax1.scatter(np.array(po),
np.array(eig2)[pos],
s=45,
marker='^',
color='black')
maxpos = np.argmax(eig2)
maxrt = rttic[np.arange(seg1[0], seg1[1])[maxpos]]
ax1.scatter(np.array([maxrt]),
np.array(eig2)[maxpos],
s=45,
marker='o',
color='blue')
ax1.set_xlim((min(x1), max(x1)))
ax1.set_ylim((0, 1.02))
plt.subplots_adjust(bottom=0.1, top=0.95, hspace=0.25)
ax1.set_ylabel('score')
ax1.set_xlabel('retention time (min)')
plt.legend(loc='upper left')
ax = plt.subplot(212)
ii = 12
y1 = x[ii, :] / norm(x[ii, :])
ax.vlines(mz,
np.zeros((len(mz), )),
y1,
color='b',
linestyles='solid',
label='R')
y2 = ms[com] / norm(ms[com])
ind = np.argsort(y2)[::-1][1:20]
yy = y2[ind]
mz1 = mz[ind]
ax.vlines(mz1,
np.zeros((len(mz1), )),
-yy,
color='r',
linestyles='solid',
label='E')
ax.plot(range(0, len(mz)), np.zeros(len(mz)), color='k')
ax.set_xlim((min(mz), 350))
ax.set_ylim((-0.25, 0.25))
ax.set_xlabel('m/z')
ax.set_ylabel('intensity')
plt.legend(loc='lower left')
plt.show()
def get_feature():
pkl_file = open('PD6--RM--')
# pkl_file = open('standardn2')
data = pickle.load(pkl_file)
msrt = data['MSRT']
fn = "F:\MARS\data_save\D6/73.CDF"
# fn = "F:\MARS\data_save\STDAND/zhi50-5vs1.CDF"
options = data['options']
ncr = netcdf_reader(fn, bmmap=False)
pw = options['pw']
w = options['w']
thre = options['thres']
rts = msrt['rt']
ms = msrt['ms']
com = 49
tz, seg, x, segx, maxpos, irr = findTZ(ncr, rts[com],
np.array(ms[com], ndmin=2), pw, w,
thre)
c, vsel, pc = gridsearch(x, maxpos, 6, irr)
c0 = ittfa(x, maxpos, 1)
c1 = ittfa(x, maxpos, 2)
c2 = ittfa(x, maxpos, 3)
c3 = ittfa(x, maxpos, 4)
c4 = ittfa(x, maxpos, 5)
c5 = ittfa(x, maxpos, 11)
#
# fig1 = plt.figure(321)
plt.subplot(322)
plt.plot(x, color='k', linewidth=1.5, alpha=0.35)
plt.subplot(324)
plt.plot(c0, 'k')
# plt.plot(c2, 'y')
# plt.plot(c3, 'c')
# plt.plot(c4, 'k')
plt.plot(c5, 'b')
plt.plot(c1, 'r')
chrom, area, heig, d = polyfitting(c, x, vsel, np.array(ms[com], ndmin=2))
ax2 = plt.subplot(326)
plt.plot(np.sum(x, axis=1))
plt.plot(chrom, 'r')
ax2.set_xlabel('scans')
com = 48
tz, seg, x, segx, maxpos, irr = findTZ(ncr, rts[com],
np.array(ms[com], ndmin=2), pw, w,
thre)
c, vsel, pc = gridsearch(x, maxpos, 6, irr)
c0 = ittfa(x, maxpos, 1)
c1 = ittfa(x, maxpos, 2)
c2 = ittfa(x, maxpos, 3)
c3 = ittfa(x, maxpos, 4)
c4 = ittfa(x, maxpos, 5)
c5 = ittfa(x, maxpos, 11)
#
# fig1 = plt.figure(321)
plt.subplot(321)
plt.plot(x, color='k', linewidth=1.5, alpha=0.35)
plt.subplot(323)
# plt.plot(c0, 'g')
plt.plot(c1, 'k')
# plt.plot(c2, 'y')
# plt.plot(c3, 'c')
plt.plot(c5, 'b')
plt.plot(c4, 'r')
chrom, area, heig, d = polyfitting(c, x, vsel, np.array(ms[com], ndmin=2))
ax1 = plt.subplot(325)
plt.plot(np.sum(x, axis=1))
plt.plot(chrom, 'r')
ax1.set_xlabel('scans')
plt.subplots_adjust(left=0.08,
right=0.95,
bottom=0.1,
top=0.9,
hspace=0.35)
plt.show()
def plot_compara2():
pkl_file = open('PD6--RM--')
data = pickle.load(pkl_file)
msrt = data['MSRT']
fn = "F:\MARS\data_save\D6/73.CDF"
options = data['options']
ncr = netcdf_reader(fn, bmmap=False)
pw = options['pw']
w = options['w']
thre = options['thres']
rts = msrt['rt']
ms = msrt['ms']
mz = msrt['mz']
com = 23
rttic = ncr.tic()['rt']
plt.figure()
ax = plt.subplot(111)
colors = ['r', 'g', 'c', 'y', 'm']
tz, seg2, x, segx, maxpos, irr = findTZ(ncr,
rts[com],
np.array(ms[com], ndmin=2),
pw,
w,
mth='RM')
# c1, vsel1 = gridsearch(x, maxpos, 6, irr)
c1 = ittfa(x, maxpos, 3)
vsel2, vsel99 = count99(x, c1)
chrom1, area, heig, d = polyfitting(c1, x, vsel2, np.array(ms[com],
ndmin=2))
print(len(vsel99))
# c2 = ittfa(x, maxpos, 2)
c2, vsel1, pc = gridsearch(x, maxpos, 6, irr)
vsel2, vsel99 = count99(x, c2)
chrom2, area, heig, d = polyfitting(c2, x, vsel2, np.array(ms[com],
ndmin=2))
print(len(vsel99))
plt.plot(rttic[range(tz[0], tz[1])],
chrom1,
color=colors[0],
linewidth=3.0,
label='com number =' + str(3))
plt.plot(rttic[range(tz[0], tz[1])],
chrom2,
color=colors[1],
linewidth=3.0,
label='com number =' + str(4))
plt.plot(rttic[range(tz[0], tz[1])],
np.sum(x, axis=1),
color='b',
linewidth=3.0,
label='raw')
ax.set_xlim((min(rttic[range(tz[0],
tz[1])]), max(rttic[range(tz[0], tz[1])])))
ax.set_xlabel('retention time (min)')
ax.set_ylabel('intensity')
plt.legend(loc='upper left')
plt.show()
# print(tt2-tt1)
# result = np.array([sics, pcs, R2, times, nums])
# np.savetxt('F:\MARS\M1.txt', result, delimiter=",", fmt="%s")
sics = []
times = []
pcs = []
R2 = []
nums = []
tt1 = time.time()
for i, m in enumerate(ms):
t1 = time.time()
tz1, seg1, x1, segx1, maxpos1, irr1 = findTZ(ncr,
rts[i],
np.array(ms[i], ndmin=2),
pw,
w,
mth='RM')
t2 = time.time()
dt1 = t2 - t1
c1, vsel, pc1 = gridsearch(x1, maxpos1, 6, irr1)
u, s, v = np.linalg.svd(x1)
xn = np.dot(np.dot(u[:, 0:pc1], np.diag(s[0:pc1])), v[0:pc1, :])
resn = x1 - xn
un = np.sum((np.power(resn, 2)))
sstn = np.sum(np.power(x1, 2))
R2.append((sstn - un) / sstn)
vsel10, vsel11, sic1, nums1 = count99(x1, c1)
chrom1, area1, heig1, r2 = polyfitting(c1, x1, vsel11,