def main():
'''
dataset_used = "german"
for i in range(3):
if i == 0:
arr = dataset_used + " " + "[1, 0, 0]"
if i == 1:
continue
if i == 2:
arr = dataset_used + " " + "[0, 0, 1]"
protected_attribute_used = attribute[i]
for l in range(5):
for m in range(4):
for n in range(4):
dataset_orig, privileged_groups, unprivileged_groups, optim_options = LoadData(dataset_used, protected_attribute_used)
algorithm_used = arr + " " + "[" + " " + str(l) + " " + str(m) + " " + str(n) + "]"
#dataset_original = copy.deepcopy(dataset_orig)
feature = comb_algorithm(l, m, n, dataset_orig, privileged_groups, unprivileged_groups, optim_options)
result = str(algorithm_used + feature)
print(result)
with open('comb_algor_german544.txt', 'a') as f:
f.write(result)
f.write("\n")
f.close()
'''
dataset_used = "german"
# if i == 0 and j == 2:
protected_attribute_used = "sex"
dataset_orig, privileged_groups, unprivileged_groups, optim_options = LoadData(
dataset_used, protected_attribute_used)
l = 2
m = 0
n = 1
feature = comb_algorithm(l, m, n, dataset_orig, privileged_groups,
unprivileged_groups, optim_options)
print(l, m, n)
print(feature)
def get_2d_array(index,chi0s,R_avg0s):
data_path = "../../2019-08-02/R_avg0_vs_chi0_scans"
loadfilepath = data_path + "/data_formatted"
datfile = data_path + "/data_formatted/input.dat"
scan = {}
loadsuf=["sigma_R0","R_eq","volFrac_0","beta","chi_0"]
savesuf=["sigma_R0","R_eq","volFrac_0","beta"]
thearray = np.zeros([len(chi0s),len(R_avg0s)],float)
for i,chi0 in enumerate(chi0s):
scan['chi_0'] = str(chi0)
ld = LoadData(name=f"scanning",scan=scan,loadfilepath=loadfilepath,
loadsuf=loadsuf,savesuf=savesuf,
datfile=datfile)
row = ld.data[:,index]
if index != 4 and index != -1:
Nrow = ld.data[:,4].astype(int)
row[np.where(Nrow <= 1)] = np.nan
thearray[i,:] = row
return thearray,ld
self.saver = tf.train.Saver()
model_path = os.path.join(backup_path, 'model.ckpt')
assert (os.path.exists(model_path + '.index'))
self.saver.restore(self.sess, model_path)
print('Successfully read model from %s' % (model_path))
#在测试集上计算准确率
accuracy_list = []
test_images = dataloader[0]
test_labels = dataloader[1]
#test_images = dataloader.data_augmenttation(dataloader.test_images, flip = False, crop = True, crop_shape = (24, 24, 3), whiten = True, noise = False)
#test_labels = dataloader.test_labels
for i in range(0, data.n_test, batch_size):
batch_images = test_images[i:i + batch_size]
batch_labels = test_labels[i:i + batch_size]
[avg_accuracy] = self.sess.run(fetches=[self.accuracy],
feed_dict={
self.images: batch_images,
self.labels: batch_labels,
self.keep_prob: 1.0
})
accuracy_list.append(avg_accuracy)
print('test precision: %.4f' % (100 * numpy.mean(accuracy_list)) + '%')
self.sess.close()
cnn = ConvNet()
data = LoadData()
#data = Corpus()
cnn.train(([data.train_images, data.train_labels
], [data.valid_images, data.valid_labels]), 'model3/')
cnn.test([data.test_images, data.test_labels], 'model3/')
for j,sigma in enumerate(sigmas):
scan['sigma_R0'] = str(sigma)
rp = ReadParams(scan=scan,datfile=datfile)
ts = rp.list_of_t_vals()
Ns = np.empty([len(ts)],float)
for i_t,t in enumerate(ts):
ld = LoadData(name=f"radius_{i_t}",scan=scan,savesuf=savesuf,
loadfilepath=loadfilepath,datfile=datfile)
Rs = ld.data[:,1]
Ns[i_t] = Rs.size
axarr.flat[i].plot(ts,Ns,markers[j],color=colors[j],
label=rf"$\sigma=\num{{{sigma:.0e}}}$")
axarr.flat[i].set_title(rf"$<\,R\,>(t=0)={R_avg0:.1f}$")
axarr.flat[i].set_yscale('log')
axarr.flat[i].set_xlabel(r"$t$")
scan['chi_0'] = chi_0(R_avg0)
for j, sigma in enumerate(sigmas):
scan['sigma_R0'] = str(sigma)
rp = ReadParams(scan=scan)
ts = rp.list_of_t_vals()
Ns = np.empty([len(ts)], float)
for i_t, t in enumerate(ts):
ld = LoadData(name=f"radius_{i_t}", scan=scan, savesuf=savesuf)
Rs = ld.data[:, 1]
Ns[i_t] = Rs.size
axarr.flat[i].plot(ts,
Ns,
markers[j],
color=colors[j],
label=rf"$\sigma=\num{{{sigma:.0e}}}$")
axarr.flat[i].set_title(rf"$<\,R\,>(t=0)={R_avg0:.1f}$")
axarr.flat[i].set_xlabel(r"$t$")
axarr.flat[i].set_ylabel(r"$N(t)$")
for j, sigma in enumerate(sigmas):
scan['sigma_R0'] = str(sigma)
rp = ReadParams(scan=scan, datfile=datfile)
ts = rp.list_of_t_vals()
R_avgts = np.empty([len(ts)], float)
sigma_Rts = np.empty([len(ts)], float)
chi_spaced = np.empty([len(ts)], float)
ldchi = LoadData(scan=scan,
savesuf=savesuf,
loadfilepath=loadfilepath,
datfile=datfile)
t_smalls = ldchi.data[:, 0]
chis = ldchi.data[:, 1]
for i_t, t in enumerate(ts):
ld = LoadData(name=f"radius_{i_t}",
scan=scan,
savesuf=savesuf,
loadfilepath=loadfilepath,
datfile=datfile)
Rs = ld.data[:, 1]
scan['sigma_R0'] = str(sigma)
rp = ReadParams(scan=scan, datfile=datfile)
ts = rp.list_of_t_vals()
print(len(ts))
fig, axarr = plt.subplots(3, len(ts) // 3)
fig.set_size_inches(width, height)
for i_t, t in enumerate(ts):
ld = LoadData(name=f"radius_{i_t}",
scan=scan,
loadfilepath=loadfilepath,
datfile=datfile)
Rs = ld.data[:, 1]
axarr.flat[i_t].hist(Rs, density=True, stacked=True)
axarr.flat[i_t].set_xlabel(r"$<R>(t)$")
axarr.flat[i_t].set_xlim(0, 20)
axarr.flat[i_t].set_ylim(0, 1)
axarr.flat[i_t].text(3, 0.5, rf"$t=\num{{{t:.1e}}}$")
fig.suptitle(rf"$<\,R\,>(t=0)={R_avg0:.1f}$" + ", " +
rf"$\sigma(t=0)=\num{{{sigma:.0e}}}$")
fig.subplots_adjust(bottom=0.08, top=0.95, left=0.08, right=0.95)
def main():
dataset_used = "adult"
for i in range(2):
if i == 0:
arr = dataset_used + " " + "[1, 0, 0]"
if i == 1:
arr = dataset_used + " " + "[0, 1, 0]"
protected_attribute_used = attribute[i]
for l in range(5):
for m in range(4):
for n in range(4):
dataset_orig, privileged_groups, unprivileged_groups, optim_options = LoadData(
dataset_used, protected_attribute_used)
algorithm_used = arr + " " + "[" + " " + str(
l) + " " + str(m) + " " + str(n) + "]"
#dataset_original = copy.deepcopy(dataset_orig)
feature = comb_algorithm(l, m, n, dataset_orig,
privileged_groups,
unprivileged_groups,
optim_options)
result = str(algorithm_used + feature)
print(result)
with open('comb_algor_adult544.txt', 'a') as f:
f.write(result)
f.write("\n")
f.close()
'''
#%%
# Imports
from loaddata import LoadData
import matplotlib.pyplot as plt
%matplotlib qt
import mglearn
from sklearn.model_selection import train_test_split
from sklearn.linear_model import LinearRegression, Ridge, Lasso, ElasticNet, LogisticRegression
from sklearn.svm import LinearSVC
from sklearn.datasets import make_blobs
import numpy as np
import pandas as pd
#%%
# Loading data
ld = LoadData()
#%%
# Linear model for wave data
X, y = ld.load_wave(n_samples=60)
X_train, X_test, y_train, y_test = train_test_split(X, y, random_state=42)
lr = LinearRegression().fit(X_train, y_train)
print(f'Train score: {lr.score(X_train,y_train)}')
print(f'Test score: {lr.score(X_test,y_test)}')
# %%
# Linear model on Boston Housing data
X, y = ld.load_boston()
X_train, X_test, y_train, y_test = train_test_split(X, y, random_state=42)
lr = LinearRegression().fit(X_train, y_train)
print(f'Train score: {lr.score(X_train,y_train)}')
def main():
dataset_used = "german"
protected_attribute_used = "sex"
arr = dataset_used + " " + "[1, 0, 0]"
data = np.zeros((80, 8))
i = 0
for l in range(5):
for m in range(4):
for n in range(4):
dataset_orig, privileged_groups, unprivileged_groups, optim_options = LoadData(
dataset_used, protected_attribute_used)
algorithm_used = arr + " " + "[" + " " + str(l) + " " + str(
m) + " " + str(n) + "]"
while (True):
feature, feature_str = comb_algorithm(
l, m, n, dataset_orig, privileged_groups,
unprivileged_groups, optim_options)
if feature[1] == 0.5:
print(feature)
continue
else:
break
print(feature)
data[i] = feature
i = i + 1
result = str(algorithm_used + feature_str)
print(result)
with open('german_sex_forAcc.txt', 'a') as f:
f.write(result)
f.write("\n")
f.close()
'''
# countries_oi = ["Italy", "Spain", "England", "France", "Netherlands", "Belgium", "Germany", "Switzerland", "Austria",
# "Hungary", "Czech Republic", "Slovakia", "Slovenia", "Poland", "Romania", "Bulgaria", "Macedonia", "Bosnia",
# "Croatia", "Serbia", "Montenegro", "Albania"] # Everything
# countries_oi = ["Netherlands","Germany","France","Belgium","Swiss German","England","Scotland","Ireland","Switzerland"]
# countries_oi_west
print("Welcome back, friend.")
while True:
inp = input(
("\nWhat do you want to do? \n(1) Extract data \n(2) Analyze Data"
"\n(8) Save/Load data \n(0) Exit \n"))
if inp == 1:
min_len = input("What is the minimum block length? (in cM)?\n")
data = LoadData(pop_path, ibd_list_path, coordinates_path, min_len,
countries_oi)
analysis = MLE_analyse(data, all_chrom=True)
elif inp == 2:
while True:
inp1 = input((
"\n(1) Regress spec. block sharing \n(2) Regress min. block sharing \n(3) Bin-Plot"
"\n(4) MLE estimation"
"\n(0) Exit \n"))
if inp1 == 1:
min_len = input("What is the minimum block length? (in cM)?\n")
max_len = input("What is the maximum block length? (in cM)?\n")
analysis.plot_ibd_spec(min_len=min_len, max_len=max_len)
elif inp1 == 2:
min_len = input("What is the minimum block length? (in cM)?\n")
analysis.plot_ibd_min(min_len, 150)
scan = {}
lines = [":", "-.", "--", "-", ":"]
for i, R_avg0 in enumerate(R_avg0s):
scan['R_avg0'] = str(R_avg0)
scan['chi_0'] = chi_0(R_avg0)
for j, sigma in enumerate(sigmas):
scan['sigma_R0'] = str(sigma)
ld = LoadData(scan=scan, savesuf=savesuf)
ts = ld.data[:, 0]
chis = ld.data[:, 1]
axarr.flat[i].set_title(rf"$<\,R\,>(t=0)={R_avg0:.1f}$")
axarr.flat[i].plot(ts,
chis,
'-',
color=colors[j],
linestyle=lines[j],
lw=4,
label=rf"$\sigma(t=0)=\num{{{sigma:.0e}}}$")
axarr.flat[i].set_xlabel(r"$t$")
scan = {}
lines = [":","-.","--","-",":"]
for i,R_avg0 in enumerate(R_avg0s):
scan['R_avg0'] = str(R_avg0)
scan['chi_0'] = chi_0(R_avg0)
for j,sigma in enumerate(sigmas):
scan['sigma_R0'] = str(sigma)
ld = LoadData(scan = scan,savesuf=savesuf,datfile=datfile,loadfilepath=loadfilepath)
ts = ld.data[:,0]
chis = ld.data[:,1]
axarr.flat[i].set_title(rf"$<\,R\,>(t=0)={R_avg0:.1f}$")
axarr.flat[i].plot(ts,chis,'-',color=colors[j],linestyle=lines[j],
lw=4,
label=rf"$\sigma(t=0)=\num{{{sigma:.0e}}}$")
axarr.flat[i].set_xlabel(r"$t$")
axarr.flat[i].set_ylabel(r"$\chi(t)$")
axarr.flat[i].legend(frameon=False,handlelength=5)
12 |rho (g/m ** 3) |1307.75 |Airtight 13 |wv (m/s) |1.03 |Wind speed 14 |max. wv (m/s) |1.75 |Maximum wind speed 15 |wd (deg) |152.3 |Wind direction in degrees 16 |rain (mm) |0.0 |Amount of the rainfall 17 |raining (s) |0.0 |Duration of the rainfall 18 |SWDR (W/m ** 2) |0.0 |Global radiation 19 |PAR (mu_mol/m**2/s) |0.0 |Photosynthetically active radiation 20 |max. PAR (mu_mol/m**2/s)|0.0 |Maximum photosynthetically active radiation 21 |Tlog (degC) |36.93 |Internal temperature of the data logger 22 |CO2 (ppm) |434.1 |CO2 concentration of the outside air """ import os import pandas as pd import matplotlib.pyplot as plt from loaddata import LoadData from parsedata import ParseData data_dir = 'data/' data = LoadData.load(data_dir=data_dir) parser = ParseData() train_data, test_data = parser.parse(data=data)
'''
Developer: vkyprmr
Filename: trees.py
Created on: 2020-09-04 at 16:45:27
'''
'''
Modified by: vkyprmr
Last modified on: 2020-09-04 at 16:45:28
'''
#%%
# Imports
from loaddata import LoadData
ld = LoadData()
from sklearn.tree import DecisionTreeClassifier
from sklearn.model_selection import train_test_split
import matplotlib.pyplot as plt
%matplotlib qt
from sklearn.tree import export_graphviz
import graphviz
import numpy as np
import pandas as pd
#%%
# Breast cancer data
X, y, fn, tn = ld.load_cancer()
X_train, X_test, y_train, y_test = train_test_split(
X, y, stratify=y, random_state=42)
tree = DecisionTreeClassifier(random_state=0)
tree.fit(X_train, y_train)
print(f"Accuracy on training set: {tree.score(X_train, y_train)}")