import matplotlib
matplotlib.use('Agg')
import numpy as np
from numpy import array
from matplotlib import pyplot
from scipy import integrate
import math
import time
from sys import path
path.append('D:pyminecleanGauss_rssi_modelimport_function')
from draw import *
import matplotlib.mlab as mlab
zhfont1 = matplotlib.font_manager.FontProperties(fname='C:WindowsFontsSTKAITI.TTF')
import sqlite3
from openpyxl import Workbook
print('ing')
def gen_rssi_model(rssi_list):
x_list = sorted(list(set(rssi_list)))
frequency_first_count, frequency_first_value, frequency_first_index, frequency_first_value_frequency, frequency_second_count, frequency_second_value, frequency_second_index, frequency_second_value_frequency = 0, 0, 0, 0, 0, 0, 0, 0
rssi_list_len = len(rssi_list)
for i in x_list:
i_value, i_count, i_index = i, rssi_list.count(i), x_list.index(i)
if i_count > frequency_first_count:
frequency_first_value, frequency_first_count, frequency_first_index, frequency_first_value_frequency = i_value, i_count, i_index, i_count / rssi_list_len
if i_count < frequency_first_count and i_count > frequency_second_count:
frequency_second_value, frequency_second_count, frequency_second_index, frequency_second_value_frequency = i_value, i_count, i_index, i_count / rssi_list_len
res_dic, gauss_rssi_model_type = {}, 1
tmp_max, tmp_min = max(frequency_first_index, frequency_second_index), min(frequency_first_index,
frequency_second_index)
frequency_first_second_middle_value, frequency_first_second_middle_count, frequency_first_second_middle_index =
x_list[tmp_min + 1], rssi_list.count((x_list[tmp_min + 1])), tmp_min + 1
len_ = len(x_list)
for i in range(0, len_, 1):
if i <= tmp_min or i >= tmp_max:
continue
i_value = x_list[i]
i_count = rssi_list.count(i_value)
i_index = i
if i_count < frequency_first_second_middle_count:
frequency_first_second_middle_value, frequency_first_second_middle_count, frequency_first_second_middle_index = i_value, i_count, i_index
if frequency_first_second_middle_value > (frequency_first_value + frequency_second_value) * 0.5:
gauss_rssi_model_type = 2
res_dic['gauss_rssi_model_type'], res_dic['frequency_first_value'], res_dic['frequency_first_count'], res_dic[
'frequency_first_index'], res_dic['frequency_second_value'], res_dic['frequency_second_count'], res_dic[
'frequency_second_index'], res_dic['frequency_first_second_middle_value'], res_dic[
'frequency_first_second_middle_count'], res_dic[
'frequency_first_second_middle_index'] = gauss_rssi_model_type, frequency_first_value, frequency_first_count, frequency_first_index, frequency_second_value, frequency_second_count, frequency_second_index, frequency_first_second_middle_value, frequency_first_second_middle_count, frequency_first_second_middle_index
return res_dic
def pdf_Normal_distribution_integrate(average_, standard_deviation, x1, x2):
f = lambda x: (np.exp(-(x - average_) ** 2 / (2 * standard_deviation ** 2))) / (
np.sqrt(2 * np.pi)) / standard_deviation
# return integrate.quad(f, x1, x2)
# TODO MODIFY try to be more precise
# return integrate.quad(f, -np.inf, x1)[0] - integrate.quad(f, -np.inf, x2)[0]
return integrate.quad(f, -np.inf, x2)[0] - integrate.quad(f, -np.inf, x1)[0]
# TODO DEL
# mlab.normpdf(len_, np_average, np_std)
def pdf_Normal_distribution_integrate_2_linear_combination(average0, standard_deviation0, average1, standard_deviation1,
x1, x2, weight0=0.5, weight1=0.5):
p0, p1 = pdf_Normal_distribution_integrate(average0, standard_deviation0, x1,
x2), pdf_Normal_distribution_integrate(average1, standard_deviation1, x1,
x2)
return weight0 * p0 + weight1 * p1
def get_list_quartern_1_3(l):
quartern_index_1 = math.ceil(len(l) / 4)
quartern_value_1 = l[quartern_index_1]
quartern_index_3 = quartern_index_1 * 3
quartern_value_3 = l[quartern_index_3]
dic_ = {}
dic_['quartern_index_1'] = quartern_index_1
dic_['quartern_value_1'] = quartern_value_1
dic_['quartern_index_3'] = quartern_index_3
dic_['quartern_value_3'] = quartern_value_3
return dic_
def draw_frequency_hist(l_, title_, xlabel='rssi', dir_='./savefig/'):
rssi_list = l_
np_std = np.std(rssi_list)
np_average = np.average(rssi_list)
data_ = array(l_)
pyplot.hist(data_, 300)
xlabel = '%s--std=%s,average=%s,sample_number=%s' % (xlabel, np_std, np_average, len(rssi_list))
pyplot.xlabel(xlabel)
pyplot.ylabel('Frequency')
localtime_ = time.strftime("%y%m%d%H%M%S", time.localtime())
title_ = '%s%s' % (title_, localtime_)
pyplot.title(title_, fontproperties=zhfont1)
dir_ = '%s%s' % (dir_, title_)
pyplot.savefig(dir_)
pyplot.close()
def draw_probability_density(rssi_list, title_, xlabel='rssi', dir_='./savefig/'):
np_std = np.std(rssi_list)
np_average = np.average(rssi_list)
x_list = sorted(list(set(rssi_list)))
len_ = len(rssi_list)
loop_ = len(x_list)
x, y = [], []
for i in range(0, loop_, 1):
val = x_list[i]
probability_density = rssi_list.count(val) / len_
x.append(val)
y.append(probability_density)
pyplot.plot(x, y)
xlabel = '%s--std=%s,average=%s,sample_number=%s' % (xlabel, np_std, np_average, len(rssi_list))
pyplot.xlabel(xlabel)
pyplot.ylabel('ProbabilityDensity')
localtime_ = time.strftime("%y%m%d%H%M%S", time.localtime())
title_ = '%s%s' % (title_, localtime_)
pyplot.title(title_, fontproperties=zhfont1)
dir_ = '%s%s' % (dir_, title_)
pyplot.savefig(dir_)
pyplot.close()
def from_db_to_res(db, sql, odd_even=0):
conn = sqlite3.connect(db)
cursor = conn.execute(sql)
res_dic, counter_ = {}, 0
for row in cursor:
counter_ += 1
if counter_ % 2 == odd_even:
continue
db_id, gather_point, mac, rssi, timestamp = row
gather_point = gather_point.replace('
', '')
if gather_point not in res_dic:
res_dic[gather_point] = {}
res_dic[gather_point]['rssi_list'] = []
res_dic[gather_point]['rssi_list'].append(rssi)
for gather_point in res_dic:
rssi_list = sorted(res_dic[gather_point]['rssi_list'])
np_std = np.std(rssi_list)
np_average = np.average(rssi_list)
res_dic[gather_point]['rssi_list_np_std'] = np_std
res_dic[gather_point]['rssi_list_np_average'] = np_average
rssi_model = gen_rssi_model(rssi_list)
res_dic[gather_point]['rssi_model'] = rssi_model
list_quartern_1_3_dic = get_list_quartern_1_3(rssi_list)
res_dic[gather_point]['quartern_index_1'], res_dic[gather_point]['quartern_value_1'], res_dic[gather_point][
'quartern_index_3'], res_dic[gather_point]['quartern_value_3'] =
list_quartern_1_3_dic['quartern_index_1'], list_quartern_1_3_dic['quartern_value_1'], list_quartern_1_3_dic[
'quartern_index_3'], list_quartern_1_3_dic['quartern_value_3']
return res_dic
db, sql = 'wifi_Tom_0814.db', 'SELECT * FROM wifi'
Tom_home_dic_even = from_db_to_res(db, sql)
Tom_home_dic_odd = from_db_to_res(db, sql, 1)
db, sql = 'wifi_beta_office_0812am.db', 'SELECT * FROM wifi WHERE belongpoint IN ("sw_office_Bata_table") '
Beta_table_dic_even = from_db_to_res(db, sql)
Beta_table_dic_odd = from_db_to_res(db, sql, 1)
k = 'sw_office_Bata_table'
rssi_list = Beta_table_dic_odd[k]['rssi_list']
title_ = '%s%s' % ('o-', k)
draw_frequency_hist_probability_density(rssi_list, title_, xlabel='rssi', dir_='./savefig/')
k = 'sw_office_Bata_table'
rssi_list = Beta_table_dic_even[k]['rssi_list']
title_ = '%s%s' % ('e-', k)
draw_frequency_hist_probability_density(rssi_list, title_, xlabel='rssi', dir_='./savefig/')
k = '下沙88栋'
rssi_list = Tom_home_dic_odd[k]['rssi_list']
title_ = '%s%s' % ('o-', k)
draw_frequency_hist_probability_density(rssi_list, title_, xlabel='rssi', dir_='./savefig/')
k = '下沙88栋'
rssi_list = Tom_home_dic_even[k]['rssi_list']
title_ = '%s%s' % ('e-', k)
draw_frequency_hist_probability_density(rssi_list, title_, xlabel='rssi', dir_='./savefig/')
report_dic = {}
def compute_relative_integrate(dic_, dic_x, type_, direction_='Tom_Beta'):
global report_dic
if direction_ not in report_dic:
report_dic[direction_] = {}
report_dic[direction_][type_] = {}
report_dic[direction_][type_]['simple_dic'] = {}
if direction_ == 'Tom_Beta':
dic_, dic_x = dic_['下沙88栋'], dic_x['sw_office_Bata_table']
elif direction_ == 'Tom_Tom':
dic_, dic_x = dic_['下沙88栋'], dic_x['下沙88栋']
elif direction_ == 'Beta_Tom':
dic_, dic_x = dic_['sw_office_Bata_table'], dic_x['下沙88栋']
elif direction_ == 'Beta_Beta':
dic_, dic_x = dic_['sw_office_Bata_table'], dic_x['sw_office_Bata_table']
average_, standard_deviation, x1, x2 = dic_['rssi_list_np_average'], dic_['rssi_list_np_std'], dic_x[
'quartern_value_1'], dic_x['quartern_value_3']
res = pdf_Normal_distribution_integrate(average_, standard_deviation, x1, x2)
simple_dic = {}
simple_dic['integrand'], simple_dic['to'], simple_dic['res'] = dic_, dic_x, res
report_dic[direction_][type_]['simple_dic'] = simple_dic
return res
# TODO MODIFY
dic_, dic_x = Tom_home_dic_even, Beta_table_dic_even
Tom_e_Beta_e_pdf_integrate = compute_relative_integrate(dic_, dic_x, 'e_e')
dic_, dic_x = Tom_home_dic_even, Beta_table_dic_odd
Tom_e_Beta_o_pdf_integrate = compute_relative_integrate(dic_, dic_x, 'e_o')
dic_, dic_x = Tom_home_dic_odd, Beta_table_dic_even
Tom_o_Beta_e_pdf_integrate = compute_relative_integrate(dic_, dic_x, 'o_e')
dic_, dic_x = Tom_home_dic_odd, Beta_table_dic_odd
Tom_o_Beta_o_pdf_integrate = compute_relative_integrate(dic_, dic_x, 'o_o')
dic_, dic_x = Tom_home_dic_odd, Tom_home_dic_odd
Tom_o_Tom_o_pdf_integrate = compute_relative_integrate(dic_, dic_x, 'o_o', 'Tom_Tom')
dic_, dic_x = Tom_home_dic_odd, Tom_home_dic_even
Tom_o_Tom_e_pdf_integrate = compute_relative_integrate(dic_, dic_x, 'o_e', 'Tom_Tom')
dic_, dic_x = Tom_home_dic_even, Tom_home_dic_even
Tom_e_Tom_e_pdf_integrate = compute_relative_integrate(dic_, dic_x, 'e_e', 'Tom_Tom')
dic_, dic_x = Tom_home_dic_even, Tom_home_dic_odd
Tom_e_Tom_o_pdf_integrate = compute_relative_integrate(dic_, dic_x, 'e_o', 'Tom_Tom')
dic_, dic_x = Beta_table_dic_even, Tom_home_dic_even
Beta_e_Tom_e_pdf_integrate = compute_relative_integrate(dic_, dic_x, 'e_e', 'Beta_Tom')
dic_, dic_x = Beta_table_dic_even, Tom_home_dic_odd
Beta_e_Tom_o_pdf_integrate = compute_relative_integrate(dic_, dic_x, 'e_o', 'Beta_Tom')
dic_, dic_x = Beta_table_dic_odd, Tom_home_dic_even
Beta_o_Tom_e_pdf_integrate = compute_relative_integrate(dic_, dic_x, 'o_e', 'Beta_Tom')
dic_, dic_x = Beta_table_dic_odd, Tom_home_dic_odd
Beta_o_Tom_o_pdf_integrate = compute_relative_integrate(dic_, dic_x, 'o_o', 'Beta_Tom')
dic_, dic_x = Beta_table_dic_odd, Beta_table_dic_odd
Beta_o_Beta_o_pdf_integrate = compute_relative_integrate(dic_, dic_x, 'o_o', 'Beta_Beta')
dic_, dic_x = Beta_table_dic_odd, Beta_table_dic_even
Beta_o_Beta_e_pdf_integrate = compute_relative_integrate(dic_, dic_x, 'o_e', 'Beta_Beta')
dic_, dic_x = Beta_table_dic_even, Beta_table_dic_even
Beta_e_Beta_e_pdf_integrate = compute_relative_integrate(dic_, dic_x, 'e_e', 'Beta_Beta')
dic_, dic_x = Beta_table_dic_even, Beta_table_dic_odd
Beta_e_Beta_o_pdf_integrate = compute_relative_integrate(dic_, dic_x, 'e_o', 'Beta_Beta')
for direction_ in report_dic:
for type_ in report_dic[direction_]:
ll = []
simple_dic = report_dic[direction_][type_]['simple_dic']
to_dic, integrand_dic = simple_dic['to'], simple_dic['integrand']
to_quartern_index_1, to_quartern_index_3, to_quartern_value_1, to_quartern_value_3, to_rssi_list_np_average, to_rssi_list_np_std =
to_dic['quartern_index_1'], to_dic['quartern_index_3'], to_dic['quartern_value_1'], to_dic[
'quartern_value_3'],
to_dic['rssi_list_np_average'], to_dic['rssi_list_np_std']
integrand_quartern_index_1, integrand_quartern_index_3, integrand_quartern_value_1, integrand_quartern_value_3, integrand_rssi_list_np_average, integrand_rssi_list_np_std =
integrand_dic['quartern_index_1'], integrand_dic['quartern_index_3'], integrand_dic['quartern_value_1'],
integrand_dic['quartern_value_3'], integrand_dic['rssi_list_np_average'], integrand_dic['rssi_list_np_std']
to_rssi_model, integrand_rssi_model = to_dic['rssi_model'], integrand_dic['rssi_model']
to_gauss_rssi_model_type, to_frequency_first_value, to_frequency_first_count, to_frequency_first_index, to_frequency_second_value, to_frequency_second_count, to_frequency_second_index, to_frequency_first_second_middle_value, to_frequency_first_second_middle_count, to_frequency_first_second_middle_index =
to_rssi_model['gauss_rssi_model_type'], to_rssi_model['frequency_first_value'],
to_rssi_model['frequency_first_count'], to_rssi_model['frequency_first_index'],
to_rssi_model['frequency_second_value'], to_rssi_model['frequency_second_count'],
to_rssi_model['frequency_second_index'], to_rssi_model['frequency_first_second_middle_value'],
to_rssi_model['frequency_first_second_middle_count'], to_rssi_model[
'frequency_first_second_middle_index']
integrand_gauss_rssi_model_type, integrand_frequency_first_value, integrand_frequency_first_count, integrand_frequency_first_index, integrand_frequency_second_value, integrand_frequency_second_count, integrand_frequency_second_index, integrand_frequency_first_second_middle_value, integrand_frequency_first_second_middle_count, integrand_frequency_first_second_middle_index =
integrand_rssi_model['gauss_rssi_model_type'], integrand_rssi_model['frequency_first_value'],
integrand_rssi_model['frequency_first_count'], integrand_rssi_model['frequency_first_index'],
integrand_rssi_model['frequency_second_value'], integrand_rssi_model['frequency_second_count'],
integrand_rssi_model['frequency_second_index'], integrand_rssi_model['frequency_first_second_middle_value'],
integrand_rssi_model['frequency_first_second_middle_count'], integrand_rssi_model[
'frequency_first_second_middle_index']
res_single = pdf_Normal_distribution_integrate(integrand_rssi_list_np_average, integrand_rssi_list_np_std,
to_quartern_value_1, to_quartern_value_3)
# 双峰模型：假设两个“子分布”均为正太分布且离散程度相同均等于全量数据的方差
res_double = pdf_Normal_distribution_integrate_2_linear_combination(integrand_frequency_first_value,
integrand_rssi_list_np_std,
integrand_frequency_second_value,
integrand_rssi_list_np_std,
to_quartern_value_1, to_quartern_value_3)
report_dic[direction_][type_]['simple_dic']['res_single'], report_dic[direction_][type_]['simple_dic'][
'res_double'] = res_single, res_double
dd = 8
# wb = Workbook()
# worksheet = wb.active
# title_ = 'direction_, type_, res_single, res_double, to_rssi_list_np_average, to_rssi_list_np_std, to_quartern_index_1, to_quartern_value_1, to_quartern_index_3, to_quartern_value_3, to_gauss_rssi_model_type, to_frequency_first_value, to_frequency_first_count, to_frequency_first_index, to_frequency_second_value, to_frequency_second_count, to_frequency_second_index, to_frequency_first_second_middle_value, to_frequency_first_second_middle_count, to_frequency_first_second_middle_index, integrand_rssi_list_np_average, integrand_rssi_list_np_std,integrand_quartern_index_1, integrand_quartern_value_1, integrand_quartern_index_3,integrand_quartern_value_3,integrand_gauss_rssi_model_type, integrand_frequency_first_value, integrand_frequency_first_count, integrand_frequency_first_index, integrand_frequency_second_value, integrand_frequency_second_count, integrand_frequency_second_index, integrand_frequency_first_second_middle_value, integrand_frequency_first_second_middle_count, integrand_frequency_first_second_middle_index'
# title_l = title_.replace(' ', '').split(',')
# worksheet.append(title_l)
# for direction_ in report_dic:
# for type_ in report_dic[direction_]:
# ll = []
# simple_dic = report_dic[direction_][type_]['simple_dic']
# to_dic, integrand_dic = simple_dic['to'], simple_dic['integrand']
# to_quartern_index_1, to_quartern_index_3, to_quartern_value_1, to_quartern_value_3, to_rssi_list_np_average, to_rssi_list_np_std =
# to_dic['quartern_index_1'], to_dic['quartern_index_3'], to_dic['quartern_value_1'], to_dic[
# 'quartern_value_3'],
# to_dic['rssi_list_np_average'], to_dic['rssi_list_np_std']
# integrand_quartern_index_1, integrand_quartern_index_3, integrand_quartern_value_1, integrand_quartern_value_3, integrand_rssi_list_np_average, integrand_rssi_list_np_std =
# integrand_dic['quartern_index_1'], integrand_dic['quartern_index_3'], integrand_dic['quartern_value_1'],
# integrand_dic['quartern_value_3'], integrand_dic['rssi_list_np_average'], integrand_dic['rssi_list_np_std']
# to_rssi_model, integrand_rssi_model = to_dic['rssi_model'], integrand_dic['rssi_model']
#
# to_gauss_rssi_model_type, to_frequency_first_value, to_frequency_first_count, to_frequency_first_index, to_frequency_second_value, to_frequency_second_count, to_frequency_second_index, to_frequency_first_second_middle_value, to_frequency_first_second_middle_count, to_frequency_first_second_middle_index =
# to_rssi_model['gauss_rssi_model_type'], to_rssi_model['frequency_first_value'],
# to_rssi_model['frequency_first_count'], to_rssi_model['frequency_first_index'],
# to_rssi_model['frequency_second_value'], to_rssi_model['frequency_second_count'],
# to_rssi_model['frequency_second_index'], to_rssi_model['frequency_first_second_middle_value'],
# to_rssi_model['frequency_first_second_middle_count'], to_rssi_model[
# 'frequency_first_second_middle_index']
#
# integrand_gauss_rssi_model_type, integrand_frequency_first_value, integrand_frequency_first_count, integrand_frequency_first_index, integrand_frequency_second_value, integrand_frequency_second_count, integrand_frequency_second_index, integrand_frequency_first_second_middle_value, integrand_frequency_first_second_middle_count, integrand_frequency_first_second_middle_index =
# integrand_rssi_model['gauss_rssi_model_type'], integrand_rssi_model['frequency_first_value'],
# integrand_rssi_model['frequency_first_count'], integrand_rssi_model['frequency_first_index'],
# integrand_rssi_model['frequency_second_value'], integrand_rssi_model['frequency_second_count'],
# integrand_rssi_model['frequency_second_index'], integrand_rssi_model['frequency_first_second_middle_value'],
# integrand_rssi_model['frequency_first_second_middle_count'], integrand_rssi_model[
# 'frequency_first_second_middle_index']
#
# res_single, res_double = report_dic[direction_][type_]['simple_dic']['res_single'],
# report_dic[direction_][type_]['simple_dic']['res_double']
#
# ll = direction_, type_, res_single, res_double, to_rssi_list_np_average, to_rssi_list_np_std, to_quartern_index_1, to_quartern_value_1, to_quartern_index_3, to_quartern_value_3, to_gauss_rssi_model_type, to_frequency_first_value, to_frequency_first_count, to_frequency_first_index, to_frequency_second_value, to_frequency_second_count, to_frequency_second_index, to_frequency_first_second_middle_value, to_frequency_first_second_middle_count, to_frequency_first_second_middle_index, integrand_rssi_list_np_average, integrand_rssi_list_np_std, integrand_quartern_index_1, integrand_quartern_value_1, integrand_quartern_index_3, integrand_quartern_value_3, integrand_gauss_rssi_model_type, integrand_frequency_first_value, integrand_frequency_first_count, integrand_frequency_first_index, integrand_frequency_second_value, integrand_frequency_second_count, integrand_frequency_second_index, integrand_frequency_first_second_middle_value, integrand_frequency_first_second_middle_count, integrand_frequency_first_second_middle_index
# worksheet.append(ll)
# file_name = '自采集数据-单双峰-概率计算结果'
# localtime_ = time.strftime("%y%m%d%H%M%S", time.localtime())
# file_name_save = '%s%s%s' % (file_name, localtime_, '.xlsx')
# wb.save(file_name_save)
#
# print('ok-finished', localtime_)
