Mysterious ship data

import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
from pylab import rcParams
rcParams['figure.figsize'] = 12, 4

import sklearn
import mysterious_ship as ms

t = np.arange(0,1000,1)
df_wind = ms.wind(t=t, u_mean=10, gust_mean=0.1,  gust_std=0.5, t_mean=300, t_std=100, components=10)
df_wind.head()

	0	1	2	3	4	5	6	7	8	9	u
0	0.021414	-0.049992	0.333796	0.769636	0.099905	0.070593	0.458019	-0.255920	0.187504	0.206486	11.841441
1	0.015030	-0.050760	0.333623	0.771459	0.097886	0.072645	0.437070	-0.250202	0.191139	0.213881	11.831772
2	0.008638	-0.051496	0.333325	0.770705	0.095630	0.074648	0.415684	-0.244394	0.194603	0.220941	11.818282
3	0.002240	-0.052201	0.332902	0.767374	0.093141	0.076601	0.393882	-0.238499	0.197891	0.227655	11.800986
4	-0.004160	-0.052874	0.332353	0.761480	0.090426	0.078503	0.371686	-0.232517	0.201000	0.234013	11.779911

fig,ax=plt.subplots()
df_wind.plot(ax=ax)

<AxesSubplot:>

data = ms.ShipData(t_max=500)
data.data_real.head()

	u	P	T	u_w	fx_wind
0.0	7.000000	7000.000000	1000	10.026627	-1.005333
0.1	6.999068	6999.068418	1000	10.027253	-1.005458
0.2	6.998137	6998.137316	1000	10.027870	-1.005582
0.3	6.997207	6997.206510	1000	10.028476	-1.005703
0.4	6.996276	6996.276250	1000	10.029074	-1.005823

for key in ['u','P']:
    fig,ax=plt.subplots()
    data.data_measure.plot(y=key, label='measured', ax=ax);
    data.data_real.plot(y=key, label='true', ax=ax);
    ax.set_ylabel(key)

from sklearn.model_selection import train_test_split
X = data.data_measure.copy()
y = X.pop('P')

test_size=0.33
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=test_size, random_state=42)

#train_index = int((1-test_size)*len(X))
#X_train = X.iloc[0:train_index]
#y_train = X.iloc[0:train_index]
#X_test = X.iloc[train_index:]
#y_test = X.iloc[train_index:]
#

for X_ in [X_train,X_test]:
    pass
    #X_['u_w**2'] = X_['u_w']**2
    #X_['u**2'] = X_['u']**2
    

from sklearn.tree import DecisionTreeRegressor

model = DecisionTreeRegressor()
model.fit(X_train, y_train)

DecisionTreeRegressor()

model.score(X_test, y_test)

0.9700538141331929