machinelearningphonons_tutorial/ML.py at master · sheriftawfikabbas/machinelearningphonons_tutorial · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
# Imports

# System libraries
import os
import warnings

# Tools

import pandas as pd  # data processing, CSV file I/O (e.g. pd.read_csv)
import numpy as np  # linear algebra
from itertools import cycle
import matplotlib.pyplot as plt
from sklearn import decomposition, datasets

# Models
from sklearn import svm, linear_model
from xgboost import XGBRegressor, plot_importance
from sklearn.linear_model import LinearRegression, TheilSenRegressor
from sklearn.ensemble import RandomForestRegressor
from sklearn.preprocessing import StandardScaler

# Utilities
from sklearn.model_selection import train_test_split
from sklearn.externals import joblib
from sklearn.metrics import r2_score, mean_squared_error, mean_absolute_error

x=pd.read_csv('https://github.com/sheriftawfikabbas/machinelearningphonons_tutorial/raw/master/X.csv',header=None)
y=pd.read_csv('https://github.com/sheriftawfikabbas/machinelearningphonons_tutorial/raw/master/y_C_v.csv',header=None)


# Visualize the C_v data. Helps in outlier detection
plt.rcParams.update({'font.size': 20})
plt.figure(figsize=(10,10))
plt.plot(np.array(y.sort_values(by=0)))
plt.savefig('Outliers', bbox_inches='tight')

# Apply scalar normalization of the input data
scaler = StandardScaler().fit(x)

# Scale the train set
x = scaler.transform(x)

# Split the input and output into the training and test sets
X_train_scaled, X_test_scaled, y_train, y_test = train_test_split(
    x, y, test_size=.2, random_state=None)

# Convert the numpy arrays into pandas dataframes
X_train_scaled=pd.DataFrame(X_train_scaled)
X_test_scaled=pd.DataFrame(X_test_scaled)
y_train=pd.DataFrame(y_train)
y_test=pd.DataFrame(y_test)

# Save the dataframes
X_train_scaled.to_csv("X_train_scaled.csv")
X_test_scaled.to_csv("X_test_scaled.csv")
y_train.to_csv("y_train.csv")
y_test.to_csv("y_test.csv")


regr_names = ['RF', 'SVM', 'XGBOOST']
regr_objects = [RandomForestRegressor(n_estimators=400, max_depth=400, random_state=0),
                svm.SVR(kernel='rbf', epsilon=0.1, verbose=True),
                XGBRegressor(objective='reg:linear', colsample_bytree=0.3, learning_rate=0.1,
                             max_depth=400, alpha=10, n_estimators=400)
                ]

for regr_choice in range(3):
    regr = regr_objects[regr_choice]
    regr_name = regr_names[regr_choice]

    regr.fit(X_train_scaled, y_train)

    y_predicted = regr.predict(X_test_scaled)

    print(mean_squared_error(y_test, y_predicted))
    print(r2_score(y_test, y_predicted))

    errors_file = open(regr_name+'_Test_Analysis.txt', 'w')
    errors_file.write(
        'RMSE\t'+str(np.sqrt(mean_squared_error(y_test, y_predicted)))+'\n')
    errors_file.write('r2\t'+str(r2_score(y_test, y_predicted))+'\n')
    errors_file.close()

    xPlot = y_test
    yPlot = y_predicted
    plt.figure(figsize=(10, 10))
    plt.plot(xPlot, yPlot, 'ro')
    plt.plot(xPlot, xPlot)
    plt.ylabel(regr_name)
    plt.xlabel('DFT')
    plt.savefig(regr_name+'_Correlation_Test', bbox_inches='tight')

    y_predicted = regr.predict(X_train_scaled)

    print(mean_squared_error(y_train, y_predicted))
    print(r2_score(y_train, y_predicted))

    errors_file = open(regr_name+'_Train_Analysis.txt', 'w')
    errors_file.write(
        'RMSE\t'+str(np.sqrt(mean_squared_error(y_train, y_predicted)))+'\n')
    errors_file.write('r2\t'+str(r2_score(y_train, y_predicted))+'\n')
    errors_file.close()

    xPlot = y_train
    yPlot = y_predicted
    plt.figure(figsize=(10, 10))
    plt.plot(xPlot, yPlot, 'ro')
    plt.plot(xPlot, xPlot)
    plt.ylabel(regr_name)
    plt.xlabel('DFT')
    plt.savefig(regr_name+'_Correlation_Train', bbox_inches='tight')

    from sklearn.externals import joblib
    joblib.dump(regr, ''+regr_name+'.pkl')