def eindexFail(self, ):
x = np.arange(0, 10, 0.5)
y1 = curve().logistic(x, a=1, b=1, c=1, d=5)
y2 = curve().logistic(x, a=1, b=1, c=2.5, d=5)
print(len(x))
print(len(y1))
print(len(y2))
dd1 = newtonDividedDifference().getDivDiffDiag(x, np.array(y1))
dd2 = newtonDividedDifference().getDivDiffDiag(x, np.array(y2))
print(dd1[-1], dd2[-1])
y_line = 0.1 * x
labels = ['logistic curve 1', 'logistic curve 2', 'y=0.1*x']
plt.plot(x,
y1,
marker='',
color='royalblue',
label=labels[0],
linewidth=3,
alpha=0.5)
plt.plot(x,
y2,
marker='',
color='r',
label=labels[1],
linewidth=3,
alpha=0.5)
plt.plot(x,
y_line,
marker='',
color='grey',
label=labels[2],
linewidth=2,
alpha=0.4)
plt.annotate("Three curves meet \nat point (0.5,0.5).",
fontsize=9,
xy=(0.52, 0.48),
xycoords='figure fraction',
xytext=(+0.52, +0.36),
bbox=dict(boxstyle="round4,pad=.5", fc="0.8"),
arrowprops=dict(
arrowstyle="fancy",
color="0.5",
shrinkB=5,
connectionstyle="arc3, rad=0.3",
))
plt.legend(fontsize=10)
plt.xlabel('Time')
plt.ylabel('Biomass')
axe = plt.gcf()
axe.set_size_inches(w=8, h=5.5)
plt.show()
def LymphoidCurve(self, ):
x = np.arange(0, 10.2, 0.5)
y = np.exp(x) / (1.5 + 1 * np.exp(1.15 * (x - 5)))
dd = newtonDividedDifference().getDivDiffDiag(x, np.array(y))
print(dd[-1])
fig, ax = plt.subplots(nrows=1, ncols=1, num=1)
plt.plot(x,
y,
marker='',
color='grey',
label='Lymphoid-typed Tissue',
linewidth=3,
alpha=0.9)
plt.hlines(y=68,
xmin=0,
xmax=10,
colors="purple",
linestyles="dashed",
linewidth=1.5,
alpha=0.6)
plt.legend(fontsize=10)
gca = plt.gca()
gca.axes.get_xaxis().set_ticks([])
gca.axes.get_yaxis().set_ticks([])
ax.spines['right'].set_visible(False)
ax.spines['top'].set_visible(False)
plt.xlabel('Time')
plt.ylabel('Growth')
axe = plt.gcf()
axe.set_size_inches(w=8, h=6.5)
plt.show()
def mouseCaseStudy(self, ):
"""
Data pooled from SM/J and 10 LG/J 535 F2 animals by James M. Cheverud.
.. @Cite:
------
James M. Cheverud. Quantitative Trait Loci for Murine Growth. Genetics, 1996.
:return:
"""
weeks = rob.IntVector(np.arange(10) + 1)
weights = rob.FloatVector([
4.83, 8.26, 12.56, 19.21, 24.80, 27.37, 29.72, 31.44, 33.71, 35.51
])
dd = newtonDividedDifference().getDivDiffCrafted(
np.array(weeks), np.array(weights))
print('Divided difference: {}'.format(dd[-1, -1]))
df = rob.DataFrame({'weeks': weeks, 'weights': weights})
rob.globalenv["y"] = df.rx2('weights')
rob.globalenv["week"] = df.rx2('weeks')
rob.globalenv["weight"] = rob.r('log((35-y)/y)')
lreg = stats.lm("weight ~ week")
a1 = lreg.rx2('coefficients')[0]
a2 = lreg.rx2('coefficients')[1]
# print(a1, a2)
t = np.array(rob.r('seq(1, length(week), 0.01)'))
z = 35 / (1 + np.exp(a1 + t * a2))
_, ax = plt.subplots(nrows=1, ncols=1, num=1)
plt.plot(t, z, color='purple', linewidth=3, alpha=0.65)
plt.plot(np.array(weeks),
np.array(weights),
color='grey',
linestyle='none',
marker='o',
alpha=0.7)
plt.xlabel("Week", fontsize=12)
plt.ylabel("Weight", fontsize=12)
plt.annotate(
"logistic function: " +
r'$ weight = \frac{35}{1+exp({2.35-0.602*week})}$',
fontsize=12,
xy=(0.47, 0.48),
xycoords='figure fraction',
)
plt.xticks(np.arange(10) + 1)
ax.spines['right'].set_visible(False)
ax.spines['top'].set_visible(False)
axe = plt.gcf()
axe.set_size_inches(w=9.5, h=6.5)
plt.show()
print(t)
def poplarBCCase(self, file_path, file_name):
to_do = np.transpose(
pd.read_table(file_path + file_name + '.csv', sep=',',
header=None))
# print(to_do)
x = np.arange(11) + 1
# print(x)
div_diff = []
for i in range(1, to_do.shape[1]):
dd = newtonDividedDifference().getDivDiffDiag(
x, np.array(to_do[i].astype(np.float)))
print(i)
div_diff.append(dd[-1])
print(dd[-1])
print(div_diff)
def calcDivDiffs(self, file_path, file_name):
to_dos = []
for i in range(len(file_name)):
f = np.transpose(
pd.read_csv(file_path + file_name[i] + '.csv',
sep=',',
header=0))
to_dos.append(f)
# print(f)
# print(pd.DataFrame(f))
div_diffs = np.zeros([10, 3])
for i in range(len(file_name)):
for j in range(to_dos[i].shape[1]):
dd = newtonDividedDifference().getDivDiffDiag(
np.arange(10) + 1, to_dos[i][j])
div_diffs[j][i] = dd[9]
print(dd[9])
pd.DataFrame(div_diffs).to_csv('./calc_dd.txt',
sep='\t',
header=None,
index=False)
return pd.DataFrame(div_diffs)