richard
/
geinf
1
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity

theoritical logLn swp2

master
tforest 2023-11-28 19:36:49 +01:00
parent 463b9e63f9
commit 2253ce5f67
1 changed files with 118 additions and 32 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

138
swp2.py
View File

@ -1,12 +1,38 @@
import matplotlib.pyplot as plt import matplotlib.pyplot as plt
import os import os
import numpy as np import numpy as np
import math
from scipy.special import gammaln
from matplotlib.backends.backend_pdf import PdfPages
def log_facto(k):
k = int(k)
if k > 1e6:
return k * np.log(k) - k + np.log(2*math.pi*k)/2
val = 0
for i in range(2, k+1):
val += np.log(i)
return val
def log_facto_1(k):
startf = 1 # start of factorial sequence
stopf = int(k+1) # end of of factorial sequence
q = gammaln(range(startf+1, stopf+1)) # n! = G(n+1)
return q[-1]
def return_x_y_from_stwp_theta_file(stwp_theta_file, breaks, mu, tgen, relative_theta_scale = False): def return_x_y_from_stwp_theta_file(stwp_theta_file, breaks, mu, tgen, relative_theta_scale = False):
with open(stwp_theta_file, "r") as swp_file: with open(stwp_theta_file, "r") as swp_file:
# Read the first line # Read the first line
line = swp_file.readline() line = swp_file.readline()
L = float(line.split()[2]) L = float(line.split()[2])
rands = swp_file.readline()
line = swp_file.readline()
# skip empty lines before SFS
while line == "\n":
line = swp_file.readline()
sfs = np.array(line.split()).astype(float)
# Process lines until the end of the file # Process lines until the end of the file
while line: while line:
# check at each line # check at each line
@ -27,7 +53,7 @@ def return_x_y_from_stwp_theta_file(stwp_theta_file, breaks, mu, tgen, relative_
#### END of parsing #### END of parsing
# quit this file if the number of dimensions is incorrect # quit this file if the number of dimensions is incorrect
if dim < breaks+1: if dim < breaks+1:
return 0,0,0 return 0,0,0,0,0
# get n, the last bin of the last group # get n, the last bin of the last group
# revert the list of groups as the most recent times correspond # revert the list of groups as the most recent times correspond
# to the closest and last leafs of the coal. tree. # to the closest and last leafs of the coal. tree.
@ -90,13 +116,19 @@ def return_x_y_from_stwp_theta_file(stwp_theta_file, breaks, mu, tgen, relative_
# # # divide by N0 # # # divide by N0
# # y[i] = y[i]/N0 # # y[i] = y[i]/N0
# # x[i] = x[i]/N0 # # x[i] = x[i]/N0
return x,y,likelihood return x,y,likelihood,sfs,L
def return_x_y_from_stwp_theta_file_as_is(stwp_theta_file, breaks, mu, tgen, relative_theta_scale = False): def return_x_y_from_stwp_theta_file_as_is(stwp_theta_file, breaks, mu, tgen, relative_theta_scale = False):
with open(stwp_theta_file, "r") as swp_file: with open(stwp_theta_file, "r") as swp_file:
# Read the first line # Read the first line
line = swp_file.readline() line = swp_file.readline()
L = float(line.split()[2]) L = float(line.split()[2])
rands = swp_file.readline()
line = swp_file.readline()
# skip empty lines before SFS
while line == "\n":
line = swp_file.readline()
sfs = np.array(line.split()).astype(float)
# Process lines until the end of the file # Process lines until the end of the file
while line: while line:
# check at each line # check at each line
@ -117,7 +149,7 @@ def return_x_y_from_stwp_theta_file_as_is(stwp_theta_file, breaks, mu, tgen, rel
#### END of parsing #### END of parsing
# quit this file if the number of dimensions is incorrect # quit this file if the number of dimensions is incorrect
if dim < breaks+1: if dim < breaks+1:
return 0,0,0 return 0,0
# get n, the last bin of the last group # get n, the last bin of the last group
# revert the list of groups as the most recent times correspond # revert the list of groups as the most recent times correspond
# to the closest and last leafs of the coal. tree. # to the closest and last leafs of the coal. tree.
@ -130,7 +162,7 @@ def return_x_y_from_stwp_theta_file_as_is(stwp_theta_file, breaks, mu, tgen, rel
groups[i] = groups[i].split(',') groups[i] = groups[i].split(',')
#print(groups[i], len(groups[i])) #print(groups[i], len(groups[i]))
thetas[i] = [float(theta_site[i]), groups[i], likelihood] thetas[i] = [float(theta_site[i]), groups[i], likelihood]
return thetas return thetas, sfs
def plot_k_epochs_thetafolder(folder_path, mu, tgen, breaks = 2, title = "Title", theta_scale = True): def plot_k_epochs_thetafolder(folder_path, mu, tgen, breaks = 2, title = "Title", theta_scale = True):
scenari = {} scenari = {}
@ -138,7 +170,7 @@ def plot_k_epochs_thetafolder(folder_path, mu, tgen, breaks = 2, title = "Title"
for file_name in os.listdir(folder_path): for file_name in os.listdir(folder_path):
if os.path.isfile(os.path.join(folder_path, file_name)): if os.path.isfile(os.path.join(folder_path, file_name)):
# Perform actions on each file # Perform actions on each file
x,y,likelihood = return_x_y_from_stwp_theta_file(folder_path+file_name, breaks = breaks, x,y,likelihood,sfs,L = return_x_y_from_stwp_theta_file(folder_path+file_name, breaks = breaks,
tgen = tgen, tgen = tgen,
mu = mu, relative_theta_scale = theta_scale) mu = mu, relative_theta_scale = theta_scale)
if x == 0 or y == 0: if x == 0 or y == 0:
@ -189,15 +221,17 @@ def plot_all_epochs_thetafolder(folder_path, mu, tgen, title = "Title", theta_sc
breaks = 0 breaks = 0
cpt +=1 cpt +=1
if os.path.isfile(os.path.join(folder_path, file_name)): if os.path.isfile(os.path.join(folder_path, file_name)):
x, y, likelihood = return_x_y_from_stwp_theta_file(folder_path+file_name, breaks = breaks, x, y, likelihood, sfs, L = return_x_y_from_stwp_theta_file(folder_path+file_name, breaks = breaks,
tgen = tgen, tgen = tgen,
mu = mu, relative_theta_scale = theta_scale) mu = mu, relative_theta_scale = theta_scale)
SFS_stored = sfs
L_stored = L
while not (x == 0 and y == 0): while not (x == 0 and y == 0):
if breaks not in epochs.keys(): if breaks not in epochs.keys():
epochs[breaks] = {} epochs[breaks] = {}
epochs[breaks][likelihood] = x,y epochs[breaks][likelihood] = x,y
breaks += 1 breaks += 1
x,y,likelihood = return_x_y_from_stwp_theta_file(folder_path+file_name, breaks = breaks, x,y,likelihood,sfs,L = return_x_y_from_stwp_theta_file(folder_path+file_name, breaks = breaks,
tgen = tgen, tgen = tgen,
mu = mu, relative_theta_scale = theta_scale) mu = mu, relative_theta_scale = theta_scale)
print("\n*******\n"+title+"\n--------\n"+"mu="+str(mu)+"\ntgen="+str(tgen)+"\nbreaks="+str(breaks)+"\n*******\n") print("\n*******\n"+title+"\n--------\n"+"mu="+str(mu)+"\ntgen="+str(tgen)+"\nbreaks="+str(breaks)+"\n*******\n")
@ -208,7 +242,7 @@ def plot_all_epochs_thetafolder(folder_path, mu, tgen, title = "Title", theta_sc
plt.figure(figsize=(5000/my_dpi, 2800/my_dpi), dpi=my_dpi) plt.figure(figsize=(5000/my_dpi, 2800/my_dpi), dpi=my_dpi)
plt.xlim(1e-3, 1) plt.xlim(1e-3, 1)
#plt.ylim(0, 10) #plt.ylim(0, 10)
#plt.yscale('log') plt.yscale('log')
plt.xscale('log') plt.xscale('log')
plt.grid(True,which="both", linestyle='--', alpha = 0.3) plt.grid(True,which="both", linestyle='--', alpha = 0.3)
brkpt_lik = [] brkpt_lik = []
@ -228,7 +262,15 @@ def plot_all_epochs_thetafolder(folder_path, mu, tgen, title = "Title", theta_sc
# divide by N0 # divide by N0
y[i] = y[i]/N0 y[i] = y[i]/N0
x[i] = x[i]/N0 x[i] = x[i]/N0
plt.plot(x, y, '-', alpha=0.75, lw=2, label = str(epoch)+' BrkPt | Lik='+greatest_likelihood) sum_theta_i = 0
print(epoch, x, y)
for i in range(2, len(y)-1):
sum_theta_i=y[i] / (i-1)
prop = []
for k in range(2, len(y)-1):
prop.append(y[k+1] / (k - 1) / sum_theta_i)
#print(epoch, prop)
plt.plot(x, y, 'o', linestyle = "-", alpha=0.75, lw=2, label = str(epoch)+' BrkPt | Lik='+greatest_likelihood)
if theta_scale: if theta_scale:
plt.xlabel("Coal. time") plt.xlabel("Coal. time")
plt.ylabel("Pop. size scaled by N0") plt.ylabel("Pop. size scaled by N0")
@ -246,35 +288,67 @@ def plot_all_epochs_thetafolder(folder_path, mu, tgen, title = "Title", theta_sc
plt.title(title) plt.title(title)
plt.savefig(title+'_b'+str(breaks)+'.pdf') plt.savefig(title+'_b'+str(breaks)+'.pdf')
# plot likelihood against nb of breakpoints # plot likelihood against nb of breakpoints
# best possible likelihood from SFS
# Segregating sites
S = sum(SFS_stored)
# number of monomorphic sites
L = L_stored
S0 = L-S
print("SFS", SFS_stored)
print("S", S, "L", L, "S0=", S0)
# compute Ln
Ln = log_facto(S+S0) - log_facto(S0) + np.log(float(S0)/(S+S0)) * S0
for xi in range(0, len(SFS_stored)):
p_i = SFS_stored[xi] / float(S+S0)
Ln += np.log(p_i) * SFS_stored[xi] - log_facto(SFS_stored[xi])
res = Ln
print(res)
# basic plot likelihood
plt.figure(figsize=(5000/my_dpi, 2800/my_dpi), dpi=my_dpi) plt.figure(figsize=(5000/my_dpi, 2800/my_dpi), dpi=my_dpi)
plt.rcParams['font.size'] = '18' plt.rcParams['font.size'] = '18'
AIC = 2*(len(brkpt_lik)+1)+2*np.array(brkpt_lik)[:, 1].astype(float) plt.plot(np.array(brkpt_lik)[:, 0], np.array(brkpt_lik)[:, 1].astype(float), 'o', linestyle = "dotted", lw=2)
plt.plot(np.array(brkpt_lik)[:, 0], AIC, 'o', linestyle = "dotted", lw=2) # plt.ylim(0,100)
plt.axhline(y=106) # plt.axhline(y=res)
plt.yscale('log') plt.yscale('log')
plt.xlabel("# breakpoints", fontsize=20) plt.xlabel("# breakpoints", fontsize=20)
plt.ylabel("$-\log\mathcal{L}$") plt.ylabel("$-\log\mathcal{L}$")
#plt.legend(loc='upper right') #plt.legend(loc='upper right')
plt.title(title) plt.title(title)
plt.savefig(title+'_Breakpts_Likelihood.pdf') plt.savefig(title+'_Breakpts_Likelihood.pdf')
# AIC
plt.figure(figsize=(5000/my_dpi, 2800/my_dpi), dpi=my_dpi)
plt.rcParams['font.size'] = '18'
AIC = 2*(len(brkpt_lik)+1)+2*np.array(brkpt_lik)[:, 1].astype(float)
plt.plot(np.array(brkpt_lik)[:, 0], AIC, 'o', linestyle = "dotted", lw=2)
# plt.axhline(y=106)
plt.yscale('log')
plt.xlabel("# breakpoints", fontsize=20)
plt.ylabel("AIC")
#plt.legend(loc='upper right')
plt.title(title)
plt.savefig(title+'_Breakpts_Likelihood_AIC.pdf')
def plot_test_theta(folder_path, mu, tgen, title = "Title", theta_scale = True, breaks_max = 6): def plot_test_theta(folder_path, mu, tgen, title = "Title", theta_scale = True, breaks_max = 5):
"""
Use theta values as is to do basic plots.
"""
cpt = 0 cpt = 0
epochs = {} epochs = {}
for file_name in os.listdir(folder_path): for file_name in os.listdir(folder_path):
cpt +=1 cpt +=1
if os.path.isfile(os.path.join(folder_path, file_name)): if os.path.isfile(os.path.join(folder_path, file_name)):
for k in range(breaks_max): for k in range(breaks_max):
thetas = return_x_y_from_stwp_theta_file_as_is(folder_path+file_name, breaks = k, thetas,sfs = return_x_y_from_stwp_theta_file_as_is(folder_path+file_name, breaks = k,
tgen = tgen, tgen = tgen,
mu = mu, relative_theta_scale = theta_scale) mu = mu, relative_theta_scale = theta_scale)
if thetas[0] == 0: if thetas == 0:
continue continue
epochs[k] = thetas epochs[k] = thetas
print("\n*******\n"+title+"\n--------\n"+"mu="+str(mu)+"\ntgen="+str(tgen)+"\nbreaks="+str(k)+"\n*******\n") print("\n*******\n"+title+"\n--------\n"+"mu="+str(mu)+"\ntgen="+str(tgen)+"\nbreaks="+str(k)+"\n*******\n")
print(cpt, "theta file(s) have been scanned.") print(cpt, "theta file(s) have been scanned.")
# intialize figure # intialize figure 1
my_dpi = 300 my_dpi = 300
plt.figure(figsize=(5000/my_dpi, 2800/my_dpi), dpi=my_dpi) plt.figure(figsize=(5000/my_dpi, 2800/my_dpi), dpi=my_dpi)
for epoch, theta in epochs.items(): for epoch, theta in epochs.items():
@ -294,7 +368,7 @@ def plot_test_theta(folder_path, mu, tgen, title = "Title", theta_scale = True,
plt.ylabel("theta") plt.ylabel("theta")
plt.legend(loc='upper right') plt.legend(loc='upper right')
plt.savefig(title+'_test'+str(k)+'.pdf') plt.savefig(title+'_test'+str(k)+'.pdf')
# fig 2 # fig 2 & 3
plt.figure(figsize=(5000/my_dpi, 2800/my_dpi), dpi=my_dpi) plt.figure(figsize=(5000/my_dpi, 2800/my_dpi), dpi=my_dpi)
for epoch, theta in epochs.items(): for epoch, theta in epochs.items():
groups = np.array(list(theta.values()), dtype=object)[:, 1].tolist() groups = np.array(list(theta.values()), dtype=object)[:, 1].tolist()
@ -308,29 +382,41 @@ def plot_test_theta(folder_path, mu, tgen, title = "Title", theta_scale = True,
N0 = y[0] N0 = y[0]
for i in range(len(y)): for i in range(len(y)):
y[i] = y[i]/N0 y[i] = y[i]/N0
#
x_2 = [] x_2 = []
T = 0 T = 0
# k allant de de 14 à 2
for i in range(len(x)): for i in range(len(x)):
x[i] = int(x[i]) x[i] = int(x[i])
#print(x[2]) # compute the times as: theta_k / (k*(k-1))
for i in range(0, len(x)): for i in range(0, len(x)):
k = x[i]
#print(k, y[k-2])
#theta_k = y[k] / (k*(k-1))
T += y[i] / (x[i]*(x[i]-1)) T += y[i] / (x[i]*(x[i]-1))
x_2.append(T) x_2.append(T)
# Plotting (fig 2)
plt.plot(x_2, y, 'o', linestyle="dotted", alpha=0.75, lw=2, label = str(epoch)+' brks')
plt.xlabel("# breaks")
plt.ylabel("theta")
plt.legend(loc='upper right')
plt.savefig(title+'_test'+str(k)+'.pdf')
# Plotting (fig 3) which is the same but log scale for x
plt.plot(x_2, y, 'o', linestyle="dotted", alpha=0.75, lw=2, label = str(epoch)+' brks') plt.plot(x_2, y, 'o', linestyle="dotted", alpha=0.75, lw=2, label = str(epoch)+' brks')
plt.xscale('log') plt.xscale('log')
plt.xlabel("# breaks") plt.xlabel("# breaks")
plt.ylabel("theta") plt.ylabel("theta")
plt.legend(loc='upper right') plt.legend(loc='upper right')
plt.savefig(title+'_test'+str(k)+'.pdf') plt.savefig(title+'_test'+str(k)+'_log.pdf')
#
def save_multi_image(filename):
pp = PdfPages(filename)
fig_nums = plt.get_fignums()
figs = [plt.figure(n) for n in fig_nums]
for fig in figs:
fig.savefig(pp, format='pdf')
pp.close()
def combined_plot(folder_path, mu, tgen, breaks, title = "Title", theta_scale = True):
plot_all_epochs_thetafolder(folder_path, mu, tgen, title, theta_scale)
plot_test_theta(folder_path, mu, tgen, title, theta_scale, breaks_max = breaks)
save_multi_image(title+"_combined.pdf")
if __name__ == "__main__": if __name__ == "__main__":
if len(sys.argv) != 4: if len(sys.argv) != 4: