def initial(trainFile, testFile, x, y, state):
data1 = pd.read_csv(trainFile)
data2 = pd.read_csv(testFile)
data1['L0005'] = data1['L0005'] - 10600
data2['L0005'] = data2['L0005'] - 10600
print(data2.head())
print("训练集数量：", len(data1))
print("测试集数量：", len(data2))
titleList = data1.columns.values.tolist()
print(titleList)
x_keys = []
for key in x:
x_keys.append(key)
x_keys.append(y)
x_keys.append(state)
for a in titleList:
if a not in x_keys:
del data1[a]
del data2[a]
print(data1.head())
print(data2.head())
trainData = data1[x_keys]
testData = data2[x_keys]
status = trainData[state].values
sentence = y + "~"
count = 0
for key in record_keys:
if count == 0:
sentence = sentence + key
else:
sentence = sentence + "+" + key
count = count + 1
print(sentence)
# mod = smf.phreg("futime ~ age + female + creatinine + "
# " + year",
# trainData, status=status, ties="efron")
mod = smf.phreg(sentence, trainData, status=status, ties="efron")
rslt = mod.fit()
print(rslt.summary())
# 得到h(t|X)=h0(t)exp(X^T*B)的协变量参数B
params = {}
i = 0
while i < len(record_keys):
params[record_keys[i]] = rslt.params[i]
i = i + 1
print(params)
return trainData, testData, params
def initial(file, x, y, state, id):
data = pd.read_csv(file)
data = data.dropna()
data["gender"] = (data["gender"] == "Male").astype(int)
data["Partner"] = (data["Partner"] == "Yes").astype(int)
data["Dependents"] = (data["Dependents"] == "Yes").astype(int)
data["PhoneService"] = (data["PhoneService"] == "Yes").astype(int)
data["MultipleLines"] = data['MultipleLines'].map({
'No phone service': 0,
'No': 1,
'Yes': 2
})
data["InternetService"] = data['InternetService'].map({
'No': 0,
'DSL': 1,
'Fiber optic': 2
})
data["TechSupport"] = data['TechSupport'].map({
'No internet service': 0,
'No': 1,
'Yes': 2
})
data["StreamingTV"] = data['StreamingTV'].map({
'No internet service': 0,
'No': 1,
'Yes': 2
})
data["Contract"] = data['Contract'].map({
'Month-to-month': 0,
'One year': 1,
'Two year': 2
})
data["PaperlessBilling"] = (data["PaperlessBilling"] == "Yes").astype(int)
data["PaymentMethod"] = data['PaymentMethod'].map({
'Electronic check':
0,
'Mailed check':
1,
'Bank transfer (automatic)':
2,
'Credit card (automatic)':
3
})
data["Churn"] = (data["Churn"] == "Yes").astype(int)
print(data.head())
print("最开始的数量：", len(data))
# 切分为训练集和测试集
titleList = data.columns.values.tolist()
print(titleList)
x_keys = []
for key in x:
x_keys.append(key)
x_keys.append(y)
x_keys.append(state)
x_keys.append(id)
y_keys = []
for a in titleList:
if a not in x_keys:
del data[a]
print(data.head())
X = data[x_keys]
Y = data[y_keys]
seed = 7
test_size = 0.4
trainData, testData, ab, cd = train_test_split(X,
Y,
test_size=test_size,
random_state=seed)
print("切分后训练集data：", len(trainData))
print("切分后测试集data：", len(testData))
status = trainData[state].values
sentence = y + "~"
count = 0
for key in record_keys:
if count == 0:
sentence = sentence + key
else:
sentence = sentence + "+" + key
count = count + 1
print(sentence)
# mod = smf.phreg("futime ~ age + female + creatinine + "
# " + year",
# trainData, status=status, ties="efron")
mod = smf.phreg(sentence, trainData, status=status, ties="efron")
rslt = mod.fit()
print(rslt.summary())
# 得到h(t|X)=h0(t)exp(X^T*B)的协变量参数B
params = {}
i = 0
while i < len(record_keys):
params[record_keys[i]] = rslt.params[i]
i = i + 1
print(params)
return trainData, testData, params
notgotit = EUhist.loc[~EUhist['user'].isin(reached)].groupby(
'user')['Seniority_days'].last().reset_index()
notgotit['reached'] = 0
dfEU = pd.concat([gotit, notgotit], axis=0)
dfEU = pd.merge(dfEU,
dta[['Id', 'user_types']],
left_on='user',
right_on='Id',
how='inner',
validate='1:1')
dfEU = pd.concat(
[dfEU, pd.get_dummies(dfEU['user_types'], prefix='type')], axis=1)
mod = smf.phreg("Seniority_days ~ 0 + type_2 + type_3",
status=dfEU['reached'].values,
data=dfEU,
ties="efron")
rslt = mod.fit()
print(rslt.summary())
fig, ax = plt.subplots()
for tp in range(1, 4):
sf = sm.SurvfuncRight(dfEU.loc[dfEU['user_types'] == tp, "Seniority_days"],
dfEU.loc[dfEU['user_types'] == tp, "reached"])
sf.plot(ax)
li = ax.get_lines()
li[1].set_visible(False) # removes crosses (not clear what they mean)
li[3].set_visible(False)
li[5].set_visible(False)
plt.legend((li[0], li[2], li[4]), ('Type 1', 'Type 2', 'Type 3'))
plt.ylim(0.8, 1)
print(sm_probit_manual.cov_params())
############################################################
# Cox-proportional
############################################################
data = sm.datasets.get_rdataset("flchain", "survival").data
del data["chapter"]
data = data.dropna()
data["lam"] = data["lambda"]
data["female"] = (data["sex"] == "F").astype(int)
data["year"] = data["sample.yr"] - min(data["sample.yr"])
status = data["death"].values
mod = smf.phreg(
"futime ~ age + female + creatinine + "
"np.sqrt(kappa) + np.sqrt(lam) + year + mgus",
data,
status=status,
ties="efron")
rslt = mod.fit()
print(rslt.summary())
############################################################
# phreg on loan data
############################################################
PATH1 = '~/Google Drive/PWBM/CECL/data'
# PATH1 = 'c:/Users/hanjh/Documents/Google Drive/PWBM/CECL/data/fannie_mae'
PATH2 = os.path.join(PATH1, 'fannie_mae', 'clean')
filename = 'noMOD_5M'
YEARS = ['2000Q4', '2001Q4', '2002Q4', '2003Q4', '2004Q4', '2005Q4']
df = pd.read_csv(os.path.join(
#!/Users/bernardroesler/anaconda3/envs/insight/bin/python3
#==============================================================================
# File: phreg_statsmodels_ex.py
# Created: 06/19/2018, 12:17
# Author: Bernie Roesler
#
"""
Description:
"""
#==============================================================================
import statsmodels.api as sm
import statsmodels.formula.api as smf
data = sm.datasets.get_rdataset("flchain", "survival").data
del data["chapter"]
data = data.dropna()
data["lam"] = data["lambda"]
data["female"] = (data["sex"] == "F").astype(int)
data["year"] = data["sample.yr"] - min(data["sample.yr"])
status = data["death"].values
mod = smf.phreg("futime ~ 0 + age + female + creatinine + "
"np.sqrt(kappa) + np.sqrt(lam) + year + mgus",
data, status=status, ties="efron")
rslt = mod.fit()
print(rslt.summary())
#==============================================================================
#==============================================================================
import pandas as pd
# from sklearn.model_selection import train_test_split
# import statsmodels.api as sm
import statsmodels.formula.api as smf
df = pd.read_csv('df_merged5.csv')
# array = df.values
# predictors = array[:, 2:-1]
# outcome = array[:, -1]
df.head()
# futime = df.iloc[:, 1]
outcome = df.iloc[:, -1]
predictors = df.iloc[:, 1:-1]
mod = smf.phreg("PERMTH_INT ~ DMPFSEQ_x + BDPEXFLR + BDPSCAN",
predictors,
status=outcome,
ties="efron")
result = mod.fit()
print(result.summary())
# X_train, X_test, y_train, y_test = train_test_split(predictors, outcome)