import numpy as np


# In[2]:


import pandas as pd


# In[3]:


import matplotlib.pyplot as plt


# In[4]:


import seaborn as sns


# # Obtendo os dados

# Lendo o arquivo UNSW-NB15_1.csv e gravando-o em um DataFrame chamado UNSW1

# In[81]:


UNSW1 = pd.read_csv('/Radhe1/CEFET/MineraçãoDados_CEFET/Projeto/The UNSW-NB15 data set description/UNSW-NB15_1.csv',dtype={"srcip":object ,},header=None)


# # Verificando o cabeçalho: head()

# In[82]:


UNSW1.head(20)


# # Usando info() e describe() no dataframe:

# In[83]:


UNSW1.info()


# In[84]:


UNSW1.describe()


# # Análise de dados exploratória e Pre-processamento
#

# In[85]:


#sns.pairplot(UNSW1,palette='bwr') #usar hue!


# In[86]:


UNSW11 = pd.read_csv('/home/govinda/Desktop/UNSW-NB15_1_ed.csv')


# In[87]:


UNSW11.describe()


# In[88]:


UNSW11['class']#a classe!


# In[89]:


#sns.pairplot(UNSW11.iloc[:,1:2],palette='bwr',hue = 'class') #usar hue!


# In[90]:


UNSW11.info()


# In[91]:


UNSW11.head()


# In[92]:


UNSW11


# In[93]:


UNSW11.dropna()


# In[94]:


print("dados 'Nan'",700001 - 22215)


# # Verificando os arquivos de treino e teste fornecidos junto ao dataset

# In[6]:


train = pd.read_csv('/home/sseg/Desktop/CEFET/UNSW_NB15_training-set.csv')


# In[7]:


train


# In[8]:


train.dropna()


# Não há dados "Nan" no conjunto de treino!

# In[10]:


train.describe()


# In[13]:


train.info()


# In[14]:


test =pd.read_csv('/home/sseg/Desktop/CEFET/UNSW_NB15_testing-set.csv')


# In[15]:


test


# In[19]:


test.dropna()


# Podemos observar que não há dados "Nan" no conjunto de teste!

# In[20]:


test.describe()


# In[21]:


test.info()


# In[23]:


test.head()


# # Tratando variáveis categóricas
#

# In[24]:


train.iloc[:,4] #state : é uma variável categórica


# In[25]:


pd.get_dummies(train['state'])


train.loc[[0],['label']]



sload =train.loc[:,'sload'] #Sload


# In[28]:


dload =train.loc[:,'dload'] #dload


# In[29]:


spkts = train.loc[:,'spkts'] #Spkts


# In[30]:


dpkts = train.loc[:,'dpkts'] #Dpkts


# In[31]:


swin = train.loc[:,'swin'] #swin


# In[32]:


dwin = train.loc[:,'dwin'] #dwin


# In[33]:


smean = train.loc[:,'smean'] #smeansz


# In[34]:


dmean = train.loc[:,'dmean'] #dmeansz


# In[35]:


sjit = train.loc[:,'sjit'] #Sjit


# In[36]:


djit = train.loc[:,'djit'] #Djit


# In[37]:


sinpkt = train.loc[:,'sinpkt'] #Sintpkt


# In[38]:


dinpkt = train.loc[:,'dinpkt'] #Dintpkt


# In[39]:


tcprtt = train.loc[:,'tcprtt'] #tcprtt


# In[40]:


synack = train.loc[:,'synack'] #synack


# In[41]:


ackdat = train.loc[:,'ackdat'] #ackdat


# In[42]:


ct_srv_src = train.loc[:,'ct_srv_src'] #ct_srv_src


# In[43]:


ct_srv_dst = train.loc[:,'ct_srv_dst'] #ct_srv_dst


# In[44]:


ct_dst_ltm = train.loc[:,'ct_dst_ltm'] #ct_dst_ltm


# In[45]:


ct_src_ltm = train.loc[:,'ct_src_ltm'] #ct_src_ltm


# In[46]:


ct_src_dport_ltm = train.loc[:,'ct_src_dport_ltm'] #ct_src_dport_ltm


# In[47]:


ct_dst_sport_ltm = train.loc[:,'ct_dst_sport_ltm'] #ct_dst_sport_ltm


# In[48]:


ct_dst_src_ltm = train.loc[:,'ct_dst_src_ltm'] #ct_dst_src_ltm


# In[49]:


classe = train.loc[:,'label']#A classe!


# # Teste:

# In[50]:


sload1 =test.loc[:,'sload'] #Sload

dload1 =test.loc[:,'dload'] #dload

spkts1 = test.loc[:,'spkts'] #Spkts

dpkts1 = test.loc[:,'dpkts'] #Dpkts

swin1 = test.loc[:,'swin'] #swin

dwin1 = test.loc[:,'dwin'] #dwin

smean1 = test.loc[:,'smean'] #smeansz

dmean1 = test.loc[:,'dmean'] #dmeansz

sjit1 = test.loc[:,'sjit'] #Sjit

djit1 = test.loc[:,'djit'] #Djit

sinpkt1 = test.loc[:,'sinpkt'] #Sintpkt

dinpkt1 = test.loc[:,'dinpkt'] #Dintpkt


tcprtt1 = test.loc[:,'tcprtt'] #tcprtt

synack1 = test.loc[:,'synack'] #synack

ackdat1 = test.loc[:,'ackdat'] #ackdat

ct_srv_src1 = test.loc[:,'ct_srv_src'] #ct_srv_src


ct_srv_dst1 = test.loc[:,'ct_srv_dst'] #ct_srv_dst

ct_dst_ltm1 = test.loc[:,'ct_dst_ltm'] #ct_dst_ltm

ct_src_ltm1 = test.loc[:,'ct_src_ltm'] #ct_src_ltm

ct_src_dport_ltm1 = test.loc[:,'ct_src_dport_ltm'] #ct_src_dport_ltm


ct_dst_sport_ltm1 = test.loc[:,'ct_dst_sport_ltm'] #ct_dst_sport_ltm


ct_dst_src_ltm1 = test.loc[:,'ct_dst_src_ltm'] #ct_dst_src_ltm


classe1 = test.loc[:,'label']#A classe!


# # #Deixando o dataframe apenas com as colunas que interessam (features):
#

# In[51]:


train = pd.concat([sload,dload,spkts,dpkts,swin,dwin,smean,dmean,sjit,djit,sinpkt,dinpkt,tcprtt,synack,ackdat,ct_srv_src,ct_srv_dst,ct_dst_ltm,ct_src_ltm,ct_src_dport_ltm,ct_dst_sport_ltm,ct_dst_src_ltm,classe],axis=1)


# In[52]:


test = pd.concat([sload1,dload1,spkts1,dpkts1,swin1,dwin1,smean1,dmean1,sjit1,djit1,sinpkt1,dinpkt1,tcprtt1,synack1,ackdat1,ct_srv_src1,ct_srv_dst1,ct_dst_ltm1,ct_src_ltm1,ct_src_dport_ltm1,ct_dst_sport_ltm1,ct_dst_src_ltm1,classe1],axis=1)


# In[53]:


train.describe()


# In[54]:


test.describe()


# # Aplicando a regressão logística

# In[57]:


from sklearn.linear_model import LogisticRegression


# In[58]:


X_train,y_train =train.drop('label',axis=1),train['label']


# In[59]:


X_test,y_test =test.drop('label',axis=1),test['label']


# In[144]:


#logmodel = LogisticRegression()


# In[60]:


import pickle  #salvar o modelo para não treinar de novo


# In[70]:


logmodel_persistente = open('logmodel_persistente.p',  'rb')


# In[71]:


pickle.load(logmodel_persistente)


# In[72]:


logmodel = logmodel_persistente


# In[73]:


#logmodel.fit(X_train,y_train) #ja salvei com o pickle


# In[74]:


predictions = logmodel.predict(X_test)


# In[166]:


import pickle  #salvar o modelo para não treinar de novo


# In[167]:


logmodel_persistente = open('logmodel_persistente.p',  'wb')


# In[168]:


pickle.dump(logmodel, logmodel_persistente)


# In[169]:


logmodel_persistente.close()


# # Avaliando o modelo

# In[173]:


from sklearn.metrics import classification_report


# In[174]:


print(classification_report(y_test,predictions))


#