In [24]:
%pylab inline
import pandas as pd
import sklearn.ensemble as sk_ensemble
import sklearn.cross_validation as sk_cv
In [21]:
def rmsle(actual, predicted):
error = np.log1p(predicted) - np.log1p(actual)
return np.sqrt(np.mean(np.square(error)))
In [2]:
train_df = pd.read_csv('../train_set_adjusted.csv')
In [3]:
print train_df.shape
print train_df.columns
In [4]:
tube = pd.read_csv('../tube_material_id_imputed_dummies_drop_ns.csv')
In [5]:
print tube.shape
print tube.columns
In [6]:
spec = pd.read_csv('../spec_dummies.csv')
In [7]:
print spec.shape
print spec.columns
In [8]:
comp = pd.read_csv('../comp_type_dummies.csv')
In [9]:
print comp.shape
print comp.columns
In [10]:
comp_type_weight = pd.read_csv('../comp_type_weight.csv')
In [11]:
print comp_type_weight.shape
print comp_type_weight.columns
In [12]:
tube_vol = pd.read_csv('../tube_volume.csv')
In [14]:
train = pd.merge(train_df, tube)
train = pd.merge(train, spec)
train = pd.merge(train, comp_type_weight)
train = pd.merge(train, tube_vol)
In [15]:
train.dtypes
Out[15]:
In [31]:
train_sub_train, train_sub_cv = sk_cv.train_test_split(train, train_size = 0.7, random_state = 0)
In [32]:
print train_sub_train.shape
print train_sub_cv.shape
In [43]:
X = train_sub_train.drop(['tube_assembly_id', 'quote_date', 'cost'], axis=1).values
Y = train_sub_train.cost
rf = sk_ensemble.RandomForestRegressor(n_estimators=100, verbose=0, random_state=0, n_jobs=4, oob_score=True)
rf = rf.fit(X, Y)
In [44]:
X_cv = train_sub_cv.drop(['tube_assembly_id', 'quote_date', 'cost'], axis=1).values
y_cv = train_sub_cv.cost
y_cv_fitted = rf.predict(X_cv)
In [45]:
rmsle(y_cv, y_cv_fitted)
Out[45]:
In [ ]: