def classifyNB(inputVec, p0Vect, p1Vect, pAbsive):
p0 = sum(inputVec * p0Vect) + log(1 - pAbsive)
p1 = sum(inputVec * p1Vect) + log(pAbsive)
if p1 > p0:
return 1
else:
return 0
def trainNB0(trainMatrix, trainCategory):
"""
朴素贝叶斯分类器训练函数
:param trainMatrix: 训练集输入向量
:param trainCategory: 每篇文档类别标签所构成的向量
:return: pAbusive：侮辱性文档的概率；
给定文档类别条件下p0:词汇表中正常单词出现的概率，p1:词汇表中侮辱性单词出现的概率
"""
numTrainDocs = len(trainMatrix)
numWords = len(trainMatrix[0])
# sum函数求和
pAbusive = sum(trainCategory) / float(numTrainDocs)
# p0Num, p1Num = zeros(numWords), zeros(numWords)
# p0Denom, p1Denom = 0.0, 0.0
# 为防止出现概率为0的情况，将所有词的出现数初始化为1，分母初始化为2
p0Num, p1Num = ones(numWords), ones(numWords)
p0Denom, p1Denom = 2.0, 2.0
for i in range(numTrainDocs):
if trainCategory[i] == 1:
p1Num += trainMatrix[i]
p1Denom += sum(trainMatrix[i])
else:
p0Num += trainMatrix[i]
p0Denom += sum(trainMatrix[i])
# p1Vect = p1Num/p1Denom
# p0Vect = p0Num/p0Denom
# 为防止下溢出，太多很小的数相乘，在python里会出现下溢出变为0
# 对每个元素除以该类别中的总词数
p1Vect = log(p1Num / p1Denom)
p0Vect = log(p0Num / p0Denom)
return p0Vect, p1Vect, pAbusive
def classifyNB(vec2Classify, p0Vec, p1Vec, pClass1):
p1 = sum(vec2Classify * p1Vec) + log(pClass1)
p0 = sum(vec2Classify * p0Vec) + log(1.0 - pClass1)
if p1 > p0:
return 1
else:
return 0
def costFunction_Regular(theta, *args):
print "现在在调用正则化的cost函数"
# print "@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@",shape(theta)
dataArr = asarray(args[0]) # args0 为特征数据 args1 为类别数据
labelArr = asarray(args[1])
m, n = shape(dataArr) # dataArr 已经加入bias
# print "记录条数:",m
theta = reshape(theta, (n, 1))
#print "costFunction_Regular中theta的值为：", theta
hx = sigmodFunction(dataArr, theta) #计算预测的类别概率值
loghx = ma.log(hx)
#print "log hx:",loghx
yhx = dot(loghx.transpose(), labelArr)
#print "yhx:",yhx
log1_hx = ma.log(1 - hx)
#print "log1_hx:",log1_hx
#print (-yhx-dot(log1_hx.transpose(),(1-labelArr)))*1.0/m
#print "lameda 值为:", args[2] # args[2]为传入的lameda参数
jtheta = (-yhx - dot(log1_hx.transpose(), (1 - labelArr))) * 1.0 / m + args[2] * 1.0 / (2 * m) * (
dot(theta.transpose(), theta) - theta[0, 0] ** 2)
#gra=getGradient(dataArr,labelArr,theta,m)
#print type(jtheta),type(gra.flatten())
#print gra
#print "###################",type(array(jtheta)[0]),array(jtheta)[0]
#print "costFunction_Regular得到的jtheta值为:", type(jtheta.flatten()[0]), jtheta, jtheta.flatten()[0]
print "&&&&&7jtheta.flatten()[0]", jtheta.flatten()[0]
return jtheta.flatten()[0]
def trainNB0(trainMatrix, trainClassMatrix):
'''
朴素贝叶斯分类器训练函数
:param trainMatrix: 训练集矩阵
:param trainClassMatrix: 包含训练集的类别矩阵
:return: p0Vect: 第0类的一个向量，其中每个维度的值表示在第0类的所有样本的所有词组中，该维度的词所占的比例, p1Vect: 意思和p0Vect类同, pAbusive: 第1类样本的数量在训练集所占的比例
'''
# 获取训练集有多少个样本
numOfTrainDocs = len(trainMatrix)
# 获取词条向量所包含的属性个数
numOfWords = len(trainMatrix[0])
pAbusive = sum(trainClassMatrix) / numOfTrainDocs
p0Num = ones(numOfWords)
p1Num = ones(numOfWords)
# 分母，表示在第0类下训练集所包含的词的总数量
p0Denom = 2.0
p1Denom = 2.0
for i in range(numOfTrainDocs):
if trainClassMatrix[i] == 1:
p1Num += trainMatrix[i]
p1Denom += sum(trainMatrix[i])
else:
p0Num += trainMatrix[i]
p0Denom += sum(trainMatrix[i])
# 这里用log是防止下溢
p0Vect = log(p0Num / p0Denom)
p1Vect = log(p1Num / p1Denom)
return p0Vect, p1Vect, pAbusive
def cost(self, x, y) -> np.ndarray:
''' Calculate the cost of output layer compare with correct data
Arguments:
x (numpy.ndarray): input data,
Shape: (total input data, input layer's total_nodes + 1)
y (numpy.ndarray): the correct output data,
Shape: (total input data, output layer's total_nodes)
Precondition:
Input layer and output layer.
'''
prediction = self.predict(x)
# Calculate Difference for y[i] = 1
a = np.multiply(y, ma.log(prediction).filled(0))
# Calculate Difference for y[i] = 0
b = np.multiply(1 - y, ma.log(1 - prediction).filled(0))
difference = a + b
total_cost = -np.sum(difference, axis=0) / len(y)
return np.sum(total_cost)
def predictNBO(p0Vec, p1Vec, pAbusive, inputVec):
p1 = sum(inputVec * p1Vec) + log(pAbusive)
p0 = sum(inputVec * p0Vec) + log(1.0 - pAbusive)
if p1 > p0:
return 1
else:
return 0
def find_an_approximation(self, function_table: dict) -> Function:
try:
SLNX = sum(log(x) for x in function_table.keys())
SLNXX = sum(log(x) * log(x) for x in function_table.keys())
SY = sum(function_table.values())
SYLNX = sum(log(x) * y for x, y in function_table.items())
n = len(function_table)
except ValueError:
return None
try:
a, b = self.solve_matrix22([[n, SLNX], [SLNX, SLNXX]], [SY, SYLNX])
if a is None:
return None
fun = lambda x: a * log(x) + b
s = sum(
(fun(x) - function_table[x])**2 for x in function_table.keys())
root_mean_square_deviation = sqrt(s / n)
f = Function(fun, f'ф = {round(a, 3)}*ln(x) {round(b, 3):+}', s,
root_mean_square_deviation)
self.print_approximation_table(function_table, f,
self.function_type)
return f
except TypeError:
return None
def average_in_flux(mag, dmag, axis=None):
flux = 10**(mag / -2.5)
dflux = np.log(10) / 2.5 * flux * dmag
avg_dflux = np.power(np.sum(np.power(dflux, -2), axis), -0.5)
avg_flux = np.sum(flux * np.power(dflux, -2), axis) * avg_dflux**2
avg_mag = -2.5 * np.log10(avg_flux)
avg_dmag = 2.5 / np.log(10) * np.divide(avg_dflux, avg_flux)
return avg_mag, avg_dmag
def entropy(array, dim=None):
if dim is None:
array = array.ravel()
dim = 0
n = ma.sum(array, dim)
array = ma.log(array) * array
sum = ma.sum(array, dim)
return (ma.log(n) - sum / n) / ma.log(2.0)
def entropy(array, dim=None):
if dim is None:
array = array.ravel()
dim = 0
n = ma.sum(array, dim)
array = ma.log(array) * array
sum = ma.sum(array, dim)
return (ma.log(n) - sum / n) / ma.log(2.0)
def average_in_flux(mag, dmag, axis=None):
flux = 10**(mag / -2.5)
dflux = np.log(10) / 2.5 * flux * dmag
avg_dflux = np.power(np.sum(np.power(dflux, -2), axis), -0.5)
avg_flux = np.sum(flux * np.power(dflux, -2), axis) * avg_dflux**2
avg_mag = -2.5 * np.log10(avg_flux)
avg_dmag = 2.5 / np.log(10) * np.divide(avg_dflux, avg_flux)
return avg_mag, avg_dmag
def computeCost(theta, X, y, lamda):
m = np.shape(X)[0]
hypo = sigmoid(X.dot(theta))
term1 = log(hypo).dot(-y)
term2 = log(1.0 - hypo).dot(1 - y)
left_hand = (term1 - term2) / m
right_hand = theta.transpose().dot(theta) * lamda / (2 * m)
return left_hand + right_hand
def log_linear_vinterp(T,P,levs):
'''
# Author Charles Doutriaux
# Version 1.1
# Expect 2D field here so there''s no reorder which I suspect to do a memory leak
# email: [email protected]
# Converts a field from sigma levels to pressure levels
# Log linear interpolation
# Input
# T : temperature on sigma levels
# P : pressure field from TOP (level 0) to BOTTOM (last level)
# levs : pressure levels to interplate to (same units as P)
# Output
# t : temperature on pressure levels (levs)
# External: Numeric'''
import numpy.ma as MA
## from numpy.oldnumeric.ma import ones,Float,greater,less,logical_and,where,equal,log,asarray,Float16
sh=P.shape
nsigma=sh[0] # Number of sigma levels
try:
nlev=len(levs) # Number of pressure levels
except:
nlev=1 # if only one level len(levs) would breaks
t=[]
for ilv in range(nlev): # loop through pressure levels
try:
lev=levs[ilv] # get value for the level
except:
lev=levs # only 1 level passed
# print ' ......... level:',lev
Pabv=MA.ones(P[0].shape,Numeric.Float)
Tabv=-Pabv # Temperature on sigma level Above
Tbel=-Pabv # Temperature on sigma level Below
Pbel=-Pabv # Pressure on sigma level Below
Pabv=-Pabv # Pressure on sigma level Above
for isg in range(1,nsigma): # loop from second sigma level to last one
## print 'Sigma level #',isg
a = MA.greater(P[isg], lev) # Where is the pressure greater than lev
b = MA.less(P[isg-1],lev) # Where is the pressure less than lev
# Now looks if the pressure level is in between the 2 sigma levels
# If yes, sets Pabv, Pbel and Tabv, Tbel
Pabv=MA.where(MA.logical_and(a,b),P[isg],Pabv) # Pressure on sigma level Above
Tabv=MA.where(MA.logical_and(a,b),T[isg],Tabv) # Temperature on sigma level Above
Pbel=MA.where(MA.logical_and(a,b),P[isg-1],Pbel) # Pressure on sigma level Below
Tbel=MA.where(MA.logical_and(a,b),T[isg-1],Tbel) # Temperature on sigma level Below
# end of for isg in range(1,nsigma)
# val=where(equal(Pbel,-1.),Pbel.missing_value,lev) # set to missing value if no data below lev if there is
tl=MA.masked_where(MA.equal(Pbel,-1.),MA.log(lev/MA.absolute(Pbel))/MA.log(Pabv/Pbel)*(Tabv-Tbel)+Tbel) # Interpolation
t.append(tl) # add a level to the output
# end of for ilv in range(nlev)
return asMA(t).astype(Numeric.Float32) # convert t to an array
def KL_Measure(i, j):
'''
计算KL散度
:return:
'''
KL1 = sum(i*(log(i/j).data))
KL2 = sum(j*(log(j/i).data))
D = (KL1 + KL2)/2
return 1/(1+ math.e ** D )
def KL_Measure(i, j):
'''
计算KL散度
:return:
'''
KL1 = sum(i * (log(i / j).data))
KL2 = sum(j * (log(j / i).data))
D = (KL1 + KL2) / 2
return 1 / (1 + math.e**D)
def muti_bin2(css, Nc, bs, nc):
ncs_1 = ma.array(
np.apply_along_axis(lambda x: np.bincount(x, minlength=Nc), 1,
css[:, bs == 1]))
ncs_0 = nc - ncs_1
N = len(bs)
nb1 = bs.sum()
nb0 = N - nb1
return (1./N * (ma.sum(ncs_1 * ma.log(1.*N*ncs_1/nc/nb1),1).filled(0) +\
ma.sum(ncs_0 * ma.log(1.*N*ncs_0/nc/nb0),1).filled(0)))
def classifyNB(vec2Classify, p0Vec, p1Vec, pClass1):
# p(w|c1)*p(c1)
# sum(vec2Classify*p1Vec)是这句话的期望?
p1 = sum(vec2Classify*p1Vec) + log(pClass1) # log(a*b) = log(a) + log(b)
# p(w1, w2, w3....|c0)*p(c0)其中每个word是相互独立的 => p(w1|c0) * p(w2|c0) * p(w3|c0) *...
# 再取对数
p0 = sum(vec2Classify*p0Vec) + log(1 - pClass1)
print("p0:%d, p1:%d" % (p0, p1))
if p1 > p0:
return 1
else:
return 0
def _compute_shannon_evenness_index(self, lct, footprint, lct_mask):
"""Compute Shannon's evenness index"""
cover_types_list = self._get_cover_types(lct)
numerator_sum = ma.masked_array(zeros(shape=lct.shape, dtype=float32),
mask=lct_mask)
for cover_type in cover_types_list:
pi = self._compute_pct_cover_type_within_footprint(lct, cover_type, footprint, lct_mask)
numerator_sum += pi*ma.filled(ma.log(pi), 0)
m = self._count_covertypes_within_window(lct, cover_types_list, footprint)
return ma.filled(-numerator_sum / ma.log(m), 0).astype(float32)
def _compute_shannon_evenness_index(self, lct, footprint, lct_mask):
"""Compute Shannon's evenness index"""
cover_types_list = self._get_cover_types(lct)
numerator_sum = ma.masked_array(zeros(shape=lct.shape, dtype=float32),
mask=lct_mask)
for cover_type in cover_types_list:
pi = self._compute_pct_cover_type_within_footprint(
lct, cover_type, footprint, lct_mask)
numerator_sum += pi * ma.filled(ma.log(pi), 0)
m = self._count_covertypes_within_window(lct, cover_types_list,
footprint)
return ma.filled(-numerator_sum / ma.log(m), 0).astype(float32)
def _zfromp_MA(P, lapse_rate, P_bott, T_bott, z_bott):
"""Altitude given pressure in a constant lapse rate layer.
The dry gas constant is used in calculations requiring the gas
constant. See the docstring for press2alt for references.
Input Arguments:
* P: Pressure [hPa].
* lapse_rate: -dT/dz [K/m] over the layer.
* P_bott: Pressure [hPa] at the base of the layer.
* T_bott: Temperature [K] at the base of the layer.
* z_bott: Geopotential altitude [m] of the base of the layer.
Output:
* Altitude [m] for each element given in the input arguments.
All input arguments can be either a scalar or an MA array. All
arguments that are MA arrays, however, are of the same size and
shape. If every input argument is a scalar, the output is a scalar.
If any of the input arguments is an MA array, the output is an MA
array of the same size and shape.
"""
import numpy as N
#jfp was import Numeric as N
import numpy.ma as MA
#jfp was import MA
from atmconst import AtmConst
const = AtmConst()
if MA.size(lapse_rate) == 1:
if MA.array(lapse_rate)[0] == 0.0:
return ( (-const.R_d * T_bott / const.g) * MA.log(P/P_bott) ) + \
z_bott
else:
exponent = (const.R_d * lapse_rate) / const.g
return ((T_bott / lapse_rate) * (1. - (P/P_bott)**exponent)) + \
z_bott
else:
exponent = (const.R_d * lapse_rate) / const.g
z = ((T_bott / lapse_rate) * (1. - (P/P_bott)**exponent)) + z_bott
z_at_0 = ( (-const.R_d * T_bott / const.g) * MA.log(P/P_bott) ) + \
z_bott
zero_lapse_mask = MA.filled(MA.where(lapse_rate == 0., 1, 0), 0)
zero_lapse_mask_indices_flat = N.nonzero(N.ravel(zero_lapse_mask))
z_flat = MA.ravel(z)
MA.put( z_flat, zero_lapse_mask_indices_flat \
, MA.take(MA.ravel(z_at_0), zero_lapse_mask_indices_flat) )
return MA.reshape(z_flat, z.shape)
def _zfromp_MA(P, lapse_rate, P_bott, T_bott, z_bott):
"""Altitude given pressure in a constant lapse rate layer.
The dry gas constant is used in calculations requiring the gas
constant. See the docstring for press2alt for references.
Input Arguments:
* P: Pressure [hPa].
* lapse_rate: -dT/dz [K/m] over the layer.
* P_bott: Pressure [hPa] at the base of the layer.
* T_bott: Temperature [K] at the base of the layer.
* z_bott: Geopotential altitude [m] of the base of the layer.
Output:
* Altitude [m] for each element given in the input arguments.
All input arguments can be either a scalar or an MA array. All
arguments that are MA arrays, however, are of the same size and
shape. If every input argument is a scalar, the output is a scalar.
If any of the input arguments is an MA array, the output is an MA
array of the same size and shape.
"""
import numpy as N
#jfp was import Numeric as N
import numpy.ma as MA
#jfp was import MA
from atmconst import AtmConst
const = AtmConst()
if MA.size(lapse_rate) == 1:
if MA.array(lapse_rate)[0] == 0.0:
return ( (-const.R_d * T_bott / const.g) * MA.log(P/P_bott) ) + \
z_bott
else:
exponent = (const.R_d * lapse_rate) / const.g
return ((T_bott / lapse_rate) * (1. - (P/P_bott)**exponent)) + \
z_bott
else:
exponent = (const.R_d * lapse_rate) / const.g
z = ((T_bott / lapse_rate) * (1. - (P / P_bott)**exponent)) + z_bott
z_at_0 = ( (-const.R_d * T_bott / const.g) * MA.log(P/P_bott) ) + \
z_bott
zero_lapse_mask = MA.filled(MA.where(lapse_rate == 0., 1, 0), 0)
zero_lapse_mask_indices_flat = N.nonzero(N.ravel(zero_lapse_mask))
z_flat = MA.ravel(z)
MA.put( z_flat, zero_lapse_mask_indices_flat \
, MA.take(MA.ravel(z_at_0), zero_lapse_mask_indices_flat) )
return MA.reshape(z_flat, z.shape)
def classifyNB(vec2Classify, p0Vec, p1Vec, pClass1):
"""
朴素贝叶斯分类函数
:param vec2Classify:
:param p0Vec:
:param p1Vec:
:param pClass1:
:return:
"""
p1 = sum(vec2Classify * p1Vec) + log(pClass1)
p0 = sum(vec2Classify * p0Vec) + log(1.0 - pClass1)
if p1 > p0:
return 1
else:
return 0
def classifyNB(vecToClassify, p0Vec, p1Vec, pClass1):
'''
利用训练好的模型进行分类
:param param:
:param p0Vec:
:param p1Vec:
:param pClass1:
:return:
'''
p1 = sum(vecToClassify * p0Vec) + log(pClass1)
p0 = sum(vecToClassify * p1Vec) + log(1 - pClass1)
if p1 > p0:
return 1
else:
return 0
def computeNMI(CG, n_G, n_C):
#CG should be a two column np array with Column 1 for Cluster and Column 2 for Class
pcg = zeros([n_G,n_C])
for line in CG:
pcg[line[0], line[1]] += 1 # column 1 is cluster and thus row
pcg = pcg/pcg.sum()
pc = pcg.sum(axis=0)
pg = pcg.sum(axis=1)
# construct pcpg, where pcpg[i,j]:= p(c=i) and p(g=j)
pcpg = zeros([n_G, n_C])
for i in range(n_G):
for j in range(n_C):
print(i,j)
pcpg[i,j] = 1/(pc[j]*pg[i])
forLog0 = multiply(pcg, pcpg)
forLog0 = ma.log(forLog0)
forLog = forLog0.filled(0)
numerator = multiply(pcg, forLog).sum()
denominatorM = -multiply(pc, log(pc)).sum() - multiply(pg, log(pg))
NMI = numerator/denominatorM.sum()
return NMI
def fit_hurdle_gamma_vector(data):
nonzero = data != 0.0
num_nonzero = np.sum(nonzero, axis = 1, keepdims = True)
num_zero = data.shape[1] - num_nonzero
insufficient_data = (num_nonzero <= SUFFICIENT_DATA_POINTS).flatten()
prob_zero = num_zero / data.shape[1]
data = ma.array(data = data, mask = ~nonzero)
# Add small number to avoid 0s in the data causing issues.
# Add small amount of noise to avoid 0s in the data causing issues
# or all values being identical causing issues.
data += np.random.uniform(1e-8, 0.2, size = data.shape)
data_mean = np.mean(data, axis = 1)
log_of_mean = np.log(data_mean)
mean_of_logs = np.mean(ma.log(data), axis = 1)
log_diff = mean_of_logs - log_of_mean
shape = 0.5 / (log_of_mean - mean_of_logs)
shape_reciprocal = 1 / shape
difference = 1
while difference > 0.000005:
numerator = log_diff + np.log(shape) - digamma(shape)
denominator = (shape ** 2) * (shape_reciprocal - polygamma(1, shape))
tmp_shape_reciprocal = shape_reciprocal + numerator / denominator
tmp_shape = 1 / tmp_shape_reciprocal
difference = np.max(np.abs(tmp_shape - shape))
shape = tmp_shape
shape_reciprocal = tmp_shape_reciprocal
scale = data_mean / shape
if np.any(np.isnan(shape)) or np.any(np.isnan(scale)):
warn("NaN shape or scale value")
return (shape.data, scale.data, prob_zero.flatten(), insufficient_data)
def gamma(self, mkt_dict_, engine_, unit_=None):
"""calculate option GAMMA with market data and engine"""
_rate, _spot, _vol, _div, _method, _param, _sign, _strike, _t = self._prepare_risk_data(
mkt_dict_, engine_)
_unit = unit_ or self.unit
if _method == EngineMethod.BS.value:
_d1 = (log(_spot / _strike) +
(_rate + _vol**2 / 2) * _t) / _vol / sqrt(_t)
return exp(-_d1**2 / 2) / sqrt(
2 * pi) / _spot / _vol / sqrt(_t) * exp(-_div * _t) * _unit
elif _method == EngineMethod.MC.value:
from utils.monte_carlo import MonteCarlo
_iteration = self._check_iter(
_param[EngineParam.MCIteration.value])
_spot = MonteCarlo.stock_price(_iteration,
isp=_spot,
rate=_rate,
div=_div,
vol=_vol,
t=_t)
_step = 0.01
_gamma = [
((max(_sign * (_s + 2 * _step - _strike), 0) -
max(_sign * (_s - _strike), 0)) -
(max(_sign * (_s - _strike), 0) -
max(_sign * (_s - 2 * _step - _strike), 0))) / (4 * _step**2)
for _s in _spot
]
return average(_gamma) * exp(-_rate * _t) * _unit
def pv(self, mkt_dict_, engine_, unit_=None):
"""calculate option PV with market data and engine"""
_rate, _spot, _vol, _div, _method, _param, _sign, _strike, _t = self._prepare_risk_data(
mkt_dict_, engine_)
_unit = unit_ or self.unit
if _method == EngineMethod.BS.value:
_d1 = (log(_spot / _strike) +
(_rate - _div + _vol**2 / 2) * _t) / _vol / sqrt(_t)
_d2 = _d1 - _vol * sqrt(_t)
return _sign * (
_spot * exp(-_div * _t) * norm.cdf(_sign * _d1) -
_strike * exp(-_rate * _t) * norm.cdf(_sign * _d2)) * _unit
elif _method == EngineMethod.MC.value:
from utils.monte_carlo import MonteCarlo
_iteration = self._check_iter(
_param[EngineParam.MCIteration.value])
_spot = MonteCarlo.stock_price(_iteration,
isp=_spot,
rate=_rate,
div=_div,
vol=_vol,
t=_t)
_price = [max(_sign * (_s - _strike), 0) for _s in _spot]
return average(_price) * exp(-_rate * _t) * _unit
def F_ideal_int(g_s, density, multiplicities):
rho_s = [density*g_i for g_i in g_s]
rho_s_ma = [ma.masked_values(rho_i/density, 0) for rho_i in rho_s]
log_s = [ma.log(rho_i_ma) for rho_i_ma in rho_s_ma]
integral = sum([m*(rho_i * ma.filled(log_i, 0) - (rho_i - density)) \
for m, rho_i, log_i in zip(multiplicities, rho_s, log_s)])
return np.sum(integral)
def preprocess(self, X, method):
"""
Preprocess the data by scaling into the range of 0-1 with bins.
"""
if method == "bucket": # scales into 0-1 range with bins
print("using the bucket prep method")
from sklearn.preprocessing import KBinsDiscretizer
est = KBinsDiscretizer(n_bins=10,
encode="ordinal",
strategy="quantile")
est.fit(X)
X_processed = est.transform(X)
X_processed /= 10 # transform from nominal values to 0-1
return X_processed
elif method == "clip": # clips the raw counts into a certain range
print("using the clip prep method")
cutoff = 1000
X_processed = np.minimum(X, cutoff) + np.sqrt(
np.maximum(X - cutoff, 0))
return X_processed
elif method == "log": # takes the log of the count
print("using the log prep method")
import numpy.ma as ma
mask = ma.log(X)
# mask logged data to replace NaN (log0) with 0
X_processed = ma.fix_invalid(mask, fill_value=0).data
return X_processed
else:
raise Exception("Incorrect preprocess method name passed!")
def F_derviative_slow(rho_s,
rho0_s,
u_s,
zr_s,
r,
x,
y,
z,
kBT,
voxel,
shape=None):
""" Derivative of 3D-RISM functional:
F'_3drism_i = kBT*np.log(g_i (r)) + u_i(r) - kBT*sum(z_ij * h_j)
return array of derivative with the shape n_s * grid
"""
if np.any(shape):
rho_s = np.reshape(rho_s, shape)
if len(zr_s.shape) == 1:
zr_s = zr_s.reshape((-1, 1))
rho_s_ma = [ma.masked_values(rho_i/rho0_i, 0, atol=1.0e-15) for rho_i, rho0_i in\
zip(rho_s, rho0_s)]
log_s = [ma.log(rho_i_ma) for rho_i_ma in rho_s_ma]
deriv = []
for i, (u_i, rho_i, log_i) in enumerate(zip(u_s, rho_s, log_s)):
# we multipy u_i by kbt as well as it is actually u_i*beta
mean_f_part = kBT * (ma.filled(log_i, -34.53877639491068) + u_i)
ex_part = 0
for rho_j, rho0_j, zr in zip(rho_s, rho0_s, zr_s.T):
delta_rho = rho_j - rho0_j
#convol = ndimage.filters.convolve(delta_rho, z_3d_ij, mode='constant')
convol = slow_convolution(delta_rho, zr, r, x, y, z, voxel)
ex_part += convol
deriv.append(mean_f_part + ex_part * kBT)
return np.array(deriv)
def F_derviative(rho_s, rho0_s, u_s, z_3ds, kBT, voxel, args, shape=None):
if np.any(shape):
rho_s = np.reshape(rho_s, shape)
rho_s_ma = [ma.masked_values(rho_i/rho0_i, 0, atol=1.0e-15) for rho_i, rho0_i in\
zip(rho_s, rho0_s)]
log_s = [ma.log(rho_i_ma) for rho_i_ma in rho_s_ma]
deriv = []
for i, (log_i, u_i, rho_i, z_3d_i) in\
enumerate(zip(log_s, u_s, rho_s, z_3ds)):
mean_f_part = kBT * (ma.filled(log_i, -34.53877639491068) + u_i)
ex_part = 0
for j, (rho_j, rho0_j,
z_3d_ij) in enumerate(zip(rho_s, rho0_s, z_3ds[i, :])):
delta_rho = rho_j - rho0_j
delta_rho_k = np.fft.fftn(delta_rho)
delta_rho_k = np.fft.fftshift(
delta_rho_k) # shift 0 freq to the middle
convol = np.real(
np.fft.ifftn(np.fft.ifftshift(delta_rho_k * z_3d_ij)))
ex_part += kBT * convol / args.multiplicities[i]
deriv_i = mean_f_part - ex_part
#fix discontinuty
deriv_i = np.where(deriv_i > 500, 500, deriv_i)
deriv.append(deriv_i)
print('median', np.median(deriv_i))
print('avg', np.average(deriv_i))
return np.array(deriv) #- 0.122324502591
def F_ideal_int(rho_s, rho0_s):
rho_s_ma = np.array([ma.masked_values(rho_i/rho0_i,0,atol=1.0e-15) for rho_i, rho0_i \
in zip(rho_s, rho0_s)])
log_s = [ma.log(rho_i_ma) for rho_i_ma in rho_s_ma]
integral = sum([rho_i * ma.filled(log_i, 0) - (rho_i - rho0_i) \
for rho_i, rho0_i, log_i in zip(rho_s, rho0_s, log_s)])
return np.sum(integral)
def __call__(self, value, clip=None):
if clip is None:
clip = self.clip
if cbook.iterable(value):
vtype = 'array'
val = ma.asarray(value).astype(np.float)
else:
vtype = 'scalar'
val = ma.array([value]).astype(np.float)
self.autoscale_None(val)
vmin, vmax = self.vmin, self.vmax
if vmin > vmax:
raise ValueError("minvalue must be less than or equal to maxvalue")
elif vmin <= 0:
raise ValueError("values must all be positive")
elif vmin == vmax:
return 0.0 * val
else:
if clip:
mask = ma.getmask(val)
val = ma.array(np.clip(val.filled(vmax), vmin, vmax),
mask=mask)
result = (ma.log(val) - np.log(vmin)) / (np.log(vmax) -
np.log(vmin))
if vtype == 'scalar':
result = result[0]
return result
def dewpoint(e):
r'''Calculate the ambient dewpoint given the vapor pressure.
Parameters
----------
e : array_like
Water vapor partial pressure in mb
Returns
-------
array_like
Dew point temperature in degrees Celsius.
See Also
--------
dewpoint_rh, saturation_vapor_pressure, vapor_pressure
Notes
-----
This function inverts the Bolton 1980 [3] formula for saturation vapor
pressure to instead calculate the temperature. This yield the following
formula for dewpoint in degrees Celsius:
.. math:: T = \frac{243.5 log(e / 6.112)}{17.67 - log(e / 6.112)}
References
----------
.. [3] Bolton, D., 1980: The Computation of Equivalent Potential
Temperature. Mon. Wea. Rev., 108, 1046-1053.
'''
val = log(e / sat_pressure_0c)
return 243.5 * val / (17.67 - val)
def zonal_avg(data,Log=False):
"""
Compute the zonal average of field on POP gx3v5 grid.
Shape of input data is expected to be either [nfoo,nlat,nlon]
or [nlat,nlon]. Log=True computes the geometric average.
Output: arrays zavg and lat
"""
print 'computing zonal average'
# get lat and lon for new regular grid
# fpin = Nio.open_file('/home/ivan/Python/data/lat_t.nc','r')
fpin = Nio.open_file('/home/emunoz/Python/mapping/model_grid/lat_t.nc','r')
lat_t = fpin.variables['lat_t'][:]
lat_t_edges = fpin.variables['lat_t_edges'][:]
fpin.close()
# fpin = Nio.open_file('/home/ivan/Python/data/gx3v5.nc','r')
fpin = Nio.open_file('/home/emunoz/Python/mapping/model_grid/gx3v5.nc','r')
lon_t = N.sort(fpin.variables['TLONG'][0,:])
ulon = N.sort(fpin.variables['ULONG'][0,:])
lon_t_edges = N.concatenate((ulon,ulon[0,N.newaxis]+360.),0)
# get gx3v5 lat and lon
tlon = fpin.variables['TLONG'][:]
tlat = fpin.variables['TLAT'][:]
fpin.close()
# compute area of cells in new regular grid
area = grid_area(lon_t_edges,lat_t_edges)
nlat = lat_t.shape[0]
nlon = lon_t.shape[0]
if data.ndim == 3:
new_data = MA.zeros((data.shape[0],nlat,nlon),dtype=float)
elif data.ndim == 2:
new_data = MA.zeros((nlat,nlon),dtype=float)
else:
print 'Check field dimensions'
sys.exit()
# geometric mean?
if Log:
work = MA.log(data)
else:
work = data
# remap data to new regular grid
for i in range(nlat):
#print 'lat = %.2f'%(lat_t[i])
for j in range(nlon):
new_data[:,i,j] = extract_loc(lon_t[j],lat_t[i],tlon,tlat,work)
# compute zonal average
if Log:
za_data = (MA.exp(MA.average(new_data,axis=-1,
weights=N.resize(area,new_data.shape))))
else:
za_data = (MA.average(new_data,axis=-1,
weights=N.resize(area,new_data.shape)))
return za_data, lat_t
def geometric_mean(array, axis=0):
'''return the geometric mean of an array removing all zero-values but
retaining total length
'''
non_zero = ma.masked_values(array, 0)
log_a = ma.log(non_zero)
return ma.exp(log_a.mean(axis=axis))
def geometric_mean(array, axis=0):
'''return the geometric mean of an array removing all zero-values but
retaining total length
'''
non_zero = ma.masked_values(array, 0)
log_a = ma.log(non_zero)
return ma.exp(log_a.mean(axis=axis))
def test_testUfuncs1(self):
# Test various functions such as sin, cos.
(x, y, a10, m1, m2, xm, ym, z, zm, xf, s) = self.d
assert_(eq(np.cos(x), cos(xm)))
assert_(eq(np.cosh(x), cosh(xm)))
assert_(eq(np.sin(x), sin(xm)))
assert_(eq(np.sinh(x), sinh(xm)))
assert_(eq(np.tan(x), tan(xm)))
assert_(eq(np.tanh(x), tanh(xm)))
with np.errstate(divide='ignore', invalid='ignore'):
assert_(eq(np.sqrt(abs(x)), sqrt(xm)))
assert_(eq(np.log(abs(x)), log(xm)))
assert_(eq(np.log10(abs(x)), log10(xm)))
assert_(eq(np.exp(x), exp(xm)))
assert_(eq(np.arcsin(z), arcsin(zm)))
assert_(eq(np.arccos(z), arccos(zm)))
assert_(eq(np.arctan(z), arctan(zm)))
assert_(eq(np.arctan2(x, y), arctan2(xm, ym)))
assert_(eq(np.absolute(x), absolute(xm)))
assert_(eq(np.equal(x, y), equal(xm, ym)))
assert_(eq(np.not_equal(x, y), not_equal(xm, ym)))
assert_(eq(np.less(x, y), less(xm, ym)))
assert_(eq(np.greater(x, y), greater(xm, ym)))
assert_(eq(np.less_equal(x, y), less_equal(xm, ym)))
assert_(eq(np.greater_equal(x, y), greater_equal(xm, ym)))
assert_(eq(np.conjugate(x), conjugate(xm)))
assert_(eq(np.concatenate((x, y)), concatenate((xm, ym))))
assert_(eq(np.concatenate((x, y)), concatenate((x, y))))
assert_(eq(np.concatenate((x, y)), concatenate((xm, y))))
assert_(eq(np.concatenate((x, y, x)), concatenate((x, ym, x))))
def test_testUfuncs1(self):
# Test various functions such as sin, cos.
(x, y, a10, m1, m2, xm, ym, z, zm, xf, s) = self.d
assert_(eq(np.cos(x), cos(xm)))
assert_(eq(np.cosh(x), cosh(xm)))
assert_(eq(np.sin(x), sin(xm)))
assert_(eq(np.sinh(x), sinh(xm)))
assert_(eq(np.tan(x), tan(xm)))
assert_(eq(np.tanh(x), tanh(xm)))
with np.errstate(divide='ignore', invalid='ignore'):
assert_(eq(np.sqrt(abs(x)), sqrt(xm)))
assert_(eq(np.log(abs(x)), log(xm)))
assert_(eq(np.log10(abs(x)), log10(xm)))
assert_(eq(np.exp(x), exp(xm)))
assert_(eq(np.arcsin(z), arcsin(zm)))
assert_(eq(np.arccos(z), arccos(zm)))
assert_(eq(np.arctan(z), arctan(zm)))
assert_(eq(np.arctan2(x, y), arctan2(xm, ym)))
assert_(eq(np.absolute(x), absolute(xm)))
assert_(eq(np.equal(x, y), equal(xm, ym)))
assert_(eq(np.not_equal(x, y), not_equal(xm, ym)))
assert_(eq(np.less(x, y), less(xm, ym)))
assert_(eq(np.greater(x, y), greater(xm, ym)))
assert_(eq(np.less_equal(x, y), less_equal(xm, ym)))
assert_(eq(np.greater_equal(x, y), greater_equal(xm, ym)))
assert_(eq(np.conjugate(x), conjugate(xm)))
assert_(eq(np.concatenate((x, y)), concatenate((xm, ym))))
assert_(eq(np.concatenate((x, y)), concatenate((x, y))))
assert_(eq(np.concatenate((x, y)), concatenate((xm, y))))
assert_(eq(np.concatenate((x, y, x)), concatenate((x, ym, x))))
def __call__(self, value, clip=None):
if clip is None:
clip = self.clip
if cbook.iterable(value):
vtype = 'array'
val = ma.asarray(value).astype(np.float)
else:
vtype = 'scalar'
val = ma.array([value]).astype(np.float)
self.autoscale_None(val)
vmin, vmax = self.vmin, self.vmax
if vmin > vmax:
raise ValueError("minvalue must be less than or equal to maxvalue")
elif vmin<=0:
raise ValueError("values must all be positive")
elif vmin==vmax:
return 0.0 * val
else:
if clip:
mask = ma.getmask(val)
val = ma.array(np.clip(val.filled(vmax), vmin, vmax),
mask=mask)
result = (ma.log(val)-np.log(vmin))/(np.log(vmax)-np.log(vmin))
if vtype == 'scalar':
result = result[0]
return result
def dewpoint(e):
r'''Calculate the ambient dewpoint given the vapor pressure.
Parameters
----------
e : array_like
Water vapor partial pressure in mb
Returns
-------
array_like
Dew point temperature in degrees Celsius.
See Also
--------
dewpoint_rh, saturation_vapor_pressure, vapor_pressure
Notes
-----
This function inverts the Bolton 1980 [3] formula for saturation vapor
pressure to instead calculate the temperature. This yield the following
formula for dewpoint in degrees Celsius:
.. math:: T = \frac{243.5 log(e / 6.112)}{17.67 - log(e / 6.112)}
References
----------
.. [3] Bolton, D., 1980: The Computation of Equivalent Potential
Temperature. Mon. Wea. Rev., 108, 1046-1053.
'''
val = log(e / sat_pressure_0c)
return 243.5 * val / (17.67 - val)
def index(request, *args, **kwags):
if request.method == 'POST':
data = ''
s = 'Invalid Input!'
form = ArForm(request.POST)
if form.is_valid():
print "VLAID DATA"
data = form.cleaned_data['article']
if len(data.strip()) < 20:
return render(request, 'classifier/index.html', {"article": ""})
dataW = fitWordArticle(data, vectorizer)
dataG = createGrammarDictionary([data])
dataF = np.concatenate((dataW[0], dataG[0]), axis=None)
dataF = ma.log(dataF/dataF.sum()*100+1).filled(0)
outside = []
outside.append(dataF)
outside = np.array(outside)
print "TESTER"
print outside.shape
ynew = clsfyModel.predict(outside)
print ynew
if ynew[0][0] > ynew[0][1]:
s = "You are a Milton Paper Writer"
else:
s = "You are a Milton Measure Writer"
return render(request, 'classifier/index.html', {"article": str(s)})
else:
return render(request, 'classifier/index.html', {"article": ""})
def transform_non_affine(self, a):
sign = np.sign(a)
masked = ma.masked_inside(a, -self.linthresh, self.linthresh, copy=False)
log = sign * self.linthresh * (self._linscale_adj + ma.log(np.abs(masked) / self.linthresh) / self._log_base)
if masked.mask.any():
return ma.where(masked.mask, a * self._linscale_adj, log)
else:
return log
def transform(self, a):
sign = np.sign(np.asarray(a))
masked = ma.masked_inside(a, -self.linthresh, self.linthresh, copy=False)
log = sign * ma.log(np.abs(masked)) / self._log_base
if masked.mask.any():
return np.asarray(ma.where(masked.mask, a * self._linadjust, log))
else:
return np.asarray(log)
def transform(self, a):
sign = np.sign(a)
masked = ma.masked_inside(a, -self.linthresh, self.linthresh, copy=False)
log = sign * self.linthresh * (1 + ma.log(np.abs(masked) / self.linthresh))
if masked.mask.any():
return ma.where(masked.mask, a, log)
else:
return log
def findEout(self, numSamples=1000):
e_out = 0
dataSamples = [self.createDataPoint() for _ in range(numSamples)]
for x, y in dataSamples:
e_out += log(1 +
exp(-1 * multiply(y, np.dot(transpose(self.w), x))))
e_out /= float(numSamples)
return e_out
def transform(self, a):
a = np.asarray(a)
sign = np.sign(a)
masked = ma.masked_inside(a, -self.linthresh, self.linthresh, copy=False)
if masked.mask.any():
log = sign * (ma.log(np.abs(masked)) / self._log_base + self._linadjust)
return np.asarray(ma.where(masked.mask, a * self._linscale, log))
else:
return sign * (np.log(np.abs(a)) / self._log_base + self._linadjust)
def ln_shifted_auto(v):
"""If 'v' has values <= 0, it is shifted in a way that min(v)=1 before doing log.
Otherwise the log is done on the original 'v'."""
vmin = ma.minimum(v)
if vmin <= 0:
values = v - vmin + 1
else:
values = v
return ma.log(values)
def trainNB1(trainMatrix, trainCategory):
numTrainDocs = len(trainMatrix)
numWord = len(trainMatrix[0])
pAbusive = sum(trainCategory) / float(numTrainDocs)
p0Num = ones(numWord)
p1Num = ones(numWord)
p0Denom = 2.0
p1Denom = 2.0
for i in range(numTrainDocs):
if trainCategory[i] == 1:
p1Num += trainMatrix[i]
p1Denom += sum(trainMatrix[i])
else:
p0Num += trainMatrix[i]
p0Denom += sum(trainMatrix[i])
p0Vect = log(p0Num / p0Denom)
p1Vect = log(p1Num / p1Denom)
return p0Vect, p1Vect, pAbusive
def perform_tf_idf(self):
words_per_doc = sum(self.a, axis=0)
docs_per_word = sum(self.a, axis=1)
rows, cols = self.a.shape
for i in range(rows):
for j in range(cols):
if words_per_doc[j] != 0:
self.a[i, j] = (self.a[i, j] / words_per_doc[j]) * log(float(cols) / docs_per_word[i])
if isinf(self.a[i, j]):
print "Infinity found!"
def corr_proba(r, ndata, ndataset=2, dof=False):
"""Probability of rejecting correlations
- **r**: Correlation coefficient
- **ndata**: Number of records use for correlations
- **ndataset**, optional: Number of datasets (1 for autocorrelations, else 2) [default: 2]
.. todo::
This must be rewritten using :mod:`scipy.stats`
"""
# Basic tests
ndata = MA.masked_equal(ndata,0,copy=0)
r = MV2.masked_where(MA.equal(MA.absolute(r),1.),r,copy=0)
# Degree of freedom
if dof:
df = ndata
else:
df = ndata-2-ndataset
# Advanced test: prevent extreme values by locally decreasing the dof
reduc = N.ones(r.shape)
z = None
while z is None or MA.count(MA.masked_greater(z,-600.)):
if z is not None:
imax = MA.argmin(z.ravel())
reduc.flat[imax] += 1
dfr = df/reduc
t = r*MV2.sqrt(dfr/((1.0-r)* (1.0+r)))
a = 0.5*dfr
b = 0.5
x = df/(dfr+t**2)
z = _gammaln(a+b)-_gammaln(a)-_gammaln(b)+a*MA.log(x)+b*MA.log(1.0-x)
# Perfom the test and format the variable
prob = MV2.masked_array(betai(a,b,x),axes=r.getAxisList())*100
prob.id = 'corr_proba' ; prob.name = prob.id
prob.long_name = 'Probability of rejection'
prob.units = '%'
return prob
def __call__(self, value, clip=None):
if clip is None:
clip = self.clip
if cbook.iterable(value):
vtype = 'array'
val = ma.asarray(value).astype(np.float)
else:
vtype = 'scalar'
val = ma.array([value]).astype(np.float)
self.autoscale_None(val)
vmin, vmax = self.vmin, self.vmax
vin, cin = self.vin, self.cin
if vmin > vmax:
raise ValueError("minvalue must be less than or equal to maxvalue")
elif vmin > 0:
raise ValueError("minvalue must be less than 0")
elif vmax < 0:
raise ValueError("maxvalue must be greater than 0")
elif vmin==vmax:
result = 0.0 * val
else:
if clip:
mask = ma.getmask(val)
val = ma.array(np.clip(val.filled(vmax), vmin, vmax),
mask=mask)
ipos = (val > vin)
ineg = (val < -vin)
izero = ~(ipos | ineg)
result = ma.empty_like(val)
result[izero] = 0.5 + cin * val[izero] / vin
result[ipos] = 0.5 + cin + (0.5 - cin) * \
(ma.log(val[ipos]) - np.log(vin)) / (np.log(vmax) - np.log(vin))
result[ineg] = 0.5 - cin - (0.5 - cin) * \
(ma.log(-val[ineg]) - np.log(vin)) / (np.log(-vmin) - np.log(vin))
result.mask = ma.getmask(val)
if vtype == 'scalar':
result = result[0]
return result
def get_diff_jpdf_with_ini(A, P0, epsilon):
"""Get joint distribution with same marginal distribution with A
and cross entropy with A is epsilon."""
t, _ = A.shape
obj_func = lambda x: dot(x, log(x))
f_eqcons = lambda x: eq_cons(x, A, epsilon)
out = fmin_slsqp(func = obj_func,
x0 = P0.reshape(-1,),
f_eqcons = f_eqcons,
bounds = [[0, 1]] *t*t,
)
return out.reshape(A.shape)
def main(tab_fname=None):
varlist = []
M = np.genfromtxt(name_iter(open(tab_fname), varlist), usemask=True, delimiter='\t')
Q = ma.MaskedArray(data=M.data, mask=(M.mask|(M.data == 0)))
Q = ma.log(Q)
# save matrix back to .tab format
fp = open(tab_fname + ".logscale.tab", "w")
for i, row in enumerate(Q):
fp.write(varlist[i] + '\t')
fp.write('\t'.join(map(tostr, row)))
fp.write('\n')
fp.close()
def dewpoint(e):
"""
Calculate the ambient dewpoint given the vapor pressure.
e : scalar or array
The water vapor partial pressure in mb.
Returns : scalar or array
The dew point temperature in degrees Celsius, with the shape
of the result being determined using numpy's broadcasting rules.
"""
val = log(e / sat_pressure_0c)
return 243.5 * val / (17.67 - val)
def costFunction_Regular(theta, *args):
# 参数顺序theta1,theta2 ,dataX_new,dataY_new
#theta=reshape(theta,())
'''
theta1=args[0] #传入初始theta的目的就是为了获得他们的维度而已
theta2=args[1]
dataX=args[2] #5000*401
m=shape(dataX)[0]
dataY=args[3] #5000*10
K=args[4] # 类别个数
lamda =args[5]
#print "dataX的维度:",shape(dataX)
#print "dataY的维度:",shape(dataY)
m1,n1=shape(theta1)
m2,n2=shape(theta2)
#print m1,n1,m1*n1
theta1=reshape(theta[0:m1*n1],(m1,n1))
theta2=reshape(theta[m1*n1:m1*n1+m2*n2],(m2,n2))
'''
theta1, theta2, dataX, K, dataY, m = getArgs(theta, args)
hx = getPredictRes(dataX, theta1, theta2) # 5000*10
#print "!@!@!@!@!@!!@!@!@!@!@!@!@@!,",shape(hx)
jtheta = 0
'''
K为类别数目，m为训练样本数
之所以这里用了一个for循环而不用向量相乘，是因为如果纯用向量相乘最终只有对角线的数据有用，如果m，n很大，这样浪费时间以及空间
由于K一般比m即训练样本的数目少很多，所以这里用向量相乘计算m个样本，外层用K个for循环，即时间复杂度为O（K），而不是O（m）
'''
for i in range(K): #dataY[:,0] 为类别1
Ak = -1 * dot(dataY[:, i].T, log(hx[:, i]))
Bk = -1 * dot((1 - dataY[:, i]).T, log(1 - hx[:, i]))
jtheta += Ak + Bk
jtheta = jtheta * 1.0 / m
#print "~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~",jtheta
reguTerm = getRegularTerm(theta1, theta2, lamda, m)
jtheta_Regular = jtheta + reguTerm
print "#################jtheta_Regular", jtheta_Regular
return jtheta_Regular
def _get_gamma_cdf(aseries, condition):
"""
Returns the CDF values for aseries.
Parameters
----------
aseries : TimeSeries
Annual series of data (one column per period)
condition : TimeSeries
Period mask.
"""
# Mask the months for which no precipitations were recorded
aseries_ = ma.masked_values(aseries, 0)
# Get the proportion of 0 precipitation for each period (MM/WW)
pzero = 1. - aseries_.count(axis=0) / aseries.count(axis=0).astype(float)
# Mask outside the reference period
aseries_._mask |= condition._data
meanrain = aseries_.mean(axis=0)
aleph = ma.log(meanrain) - ma.log(aseries_).mean(axis=0)
alpha = (1. + ma.sqrt(1.+4./3*aleph)) / (4.*aleph)
beta = meanrain/alpha
# Get the Gamma CDF (per month)
gcdf = pzero + (1.-pzero) * ssd.gamma.cdf(aseries,alpha,scale=beta)
return gcdf
def geoMean(array):
'''
Generate the geometric mean of a list or array,
removing all zero-values but retaining total length
'''
if isinstance(array, pandas.core.frame.DataFrame):
array = array.as_matrix()
else:
pass
non_zero = ma.masked_values(array,
0)
log_a = ma.log(non_zero)
geom_mean = ma.exp(log_a.mean())
return geom_mean