In [2]:

    

#Import the dependencies 

import pandas as pd
import numpy as np
from matplotlib import pyplot as plt
from sklearn import svm
from sklearn.model_selection import train_test_split


#Import the data 
train_data = pd.read_csv("dataset/train.csv")

#Clean the data
train_data.drop('idx',axis=1,inplace=True)
train_data_phase=[]
for i in xrange(1,8):
    train_ph = train_data[train_data["PH"] == i]
    train_ph.drop("PH",axis=1,inplace=True)
    train_data_phase.append(train_ph)


    

    




/usr/local/lib/python2.7/site-packages/ipykernel/__main__.py:18: SettingWithCopyWarning: 
A value is trying to be set on a copy of a slice from a DataFrame

See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/indexing.html#indexing-view-versus-copy

In [3]:

    

#Split the datasets
X_train=[]
X_test = []
y_train=[]
y_test = []

for i in xrange(0,7):
    train,test=train_test_split(train_data_phase[i], test_size = 0.20)
    y_train.append(train.ix[:,train.columns == "FF"])
    y_test.append(test.ix[:,test.columns == "FF"])
    X_train.append(train.ix[:, train.columns != "FF"])
    X_test.append(test.ix[:,test.columns != "FF"])


    

In [4]:

    

#Initialise the regressors
list_svr = []
for i in xrange(0,7):
    list_svr.append(svm.SVR())
    list_svr[i].fit(X_train[i],y_train[i])


    

    




/usr/local/lib/python2.7/site-packages/sklearn/utils/validation.py:526: DataConversionWarning: A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().
  y = column_or_1d(y, warn=True)

In [5]:

    

#Calculate the test predictions
y_pred = []
for i in xrange(0,7):
    y_pred.append(list_svr[i].predict(X_test[i]))


    

In [6]:

    

#Calculate the accuracy
from sklearn.metrics import mean_squared_error
error = 0
for i in xrange(0,7):
    error = error + mean_squared_error(y_test[i],y_pred[i])
error = error / 7.0
print error**(1./3.)  #Just a crude metric to visualise the accuracy, has no physical meaning


    

    




86.0017312706

In [131]:

    

# Get in the test data and storing the order
test_data = pd.read_csv("dataset/test.csv")
print test_data.head()
#cleaning the data and splitting across each PH
test_data.drop('idx',axis=1,inplace=True)
test_data_phase = []
for i in xrange(1,8):
    test = test_data[test_data["PH"]==i]
    test_data_phase.append(test)
    test_data_phase[i-1].drop('PH',axis=1,inplace=True)


    

    




   idx        ABRK     ELEV_1     ELEV_2  EVNT  FADF  FADS  FGC3  FIRE_1  \
0    0  119.983559  19.044380  56.936630     1    15    15   120       0   
1    1  119.983559  19.044380  56.957088     1    15    15   120       0   
2    2  119.983559  19.044380  56.957088     1    15    15   120       0   
3    3  119.983559  19.064838  56.936630     1    15    15   120       0   
4    4  119.983559  19.064838  56.916172     1    15    15   120       0   

   FIRE_2 ...   BPGR_1     BPGR_2     BPYR_1  BPYR_2  CALT   CASS       CRSS  \
0       0 ...        0  29.296949  53.711075       0     0  135.0 -29.091582   
1       0 ...        0  29.296949  53.711075       0     0  135.0 -29.091582   
2       0 ...        0  24.414124  48.828247       0     0  135.0 -29.091582   
3       0 ...        0  29.296949  48.828247       0     0  135.0 -29.091582   
4       0 ...        0  29.296949  48.828247       0     0  135.0 -29.091582   

   DFGS   DWPT  EAI  
0     1  75024    0  
1     1  75024    0  
2     1  75024    0  
3     1  75024    0  
4     1  75024    0  

[5 rows x 218 columns]






    




/usr/local/lib/python2.7/site-packages/ipykernel/__main__.py:10: SettingWithCopyWarning: 
A value is trying to be set on a copy of a slice from a DataFrame

See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/indexing.html#indexing-view-versus-copy

In [132]:

    

assert sum(test_data_phase[i].shape[0] for i in range(0,7)) == test_data.shape[0]


    

In [133]:

    

#Predict using the individual regressors
y_pred_test = []
for i in xrange(0,7):
    y_pred_test.append(list_svr[i].predict(test_data_phase[i]))


    

In [134]:

    

#write the submission
submission = np.empty([test_data.shape[0]],dtype=float)
for i in xrange(1,8):
    submission[test_data[test_data["PH"]==i].index] = y_pred_test[i-1]


    

In [151]:

    

#Save it to text file
np.savetxt("out.txt",np.ceil(submission),fmt="%d")