GaussianProcesses/data_plotter.py at master · gdroguski/GaussianProcesses · 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
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
import matplotlib.pyplot as plt
import matplotlib.dates as mdates
import numpy as np
import data_handler
import gpr_wrapper


class Plotter:
    __company_name = None
    __company_handler = None
    __prices_data = None
    __quarters = None
    __years = None
    __max_days = None
    __quarter_length = None
    __gpr = None

    def __init__(self, company_name: str):
        self.__company_name = company_name
        self.__company_handler = data_handler.CsvHandler(company_name)
        self.__prices_data = self.__company_handler.get_equal_length_prices()
        self.__quarters = self.__company_handler.quarters
        self.__years = self.__company_handler.years
        self.__max_days = self.__company_handler.max_days
        self.__quarter_length = int(self.__max_days / 4)
        self.__gpr = gpr_wrapper.Wrapper(company_name)

    def show_preprocessed_price(self, year: int):
        self.show_preprocessed_prices(start_year=year, end_year=year)

    def show_preprocessed_prices(self, start_year: int, end_year: int):
        self.__validate_dates(start_year=start_year, end_year=end_year)

        fig = plt.figure(num=self.__company_name + ' normalized prices')
        ax = plt.gca()
        fig.set_size_inches(12, 6)
        y_low, y_high = 0, 0
        for year in range(start_year, end_year + 1):
            y = self.__prices_data[year]
            if y_low >= min(y):
                y_low = min(y)
            if y_high <= max(y):
                y_high = max(y)
            x = np.linspace(0, len(y), len(y))
            plt.plot(x, y, alpha=.95, label=year)
            plt.legend(bbox_to_anchor=(1.01, 1), loc=2, borderaxespad=0.)

        y_max = max(abs(y_low) - 1, abs(y_high) + 1)
        x_min, x_max = -10, self.__max_days + 10
        ax.set_ylim(bottom=-y_max, top=y_max)
        ax.set_xlim(left=x_min, right=x_max)

        for i in range(0, 5):
            plt.vlines(x=(self.__max_days / 4) * i, ymin=-y_max, ymax=y_max, color='black', linestyles='--', alpha=.6,
                       zorder=-1)
            if i < 4:
                ax.text((self.__max_days / 4) * i + self.__max_days / 8 - 5, y_max - 0.5, self.__quarters[i],
                        fontsize=12)
        plt.hlines(y=0, xmin=x_min, xmax=x_max, color='black', linestyles='--', alpha=.6, zorder=-1)

        plt.grid(True, alpha=.25)
        plt.title(self.__company_name)
        plt.xlabel('Days')
        plt.ylabel('Normalized price')

        plt.tight_layout()

        fname = '{}_{}_{}_normalized_prices.png'.format(self.__company_name, start_year, end_year)
        fig.savefig(fname, dpi=fig.dpi)
        plt.clf()

    def show_gp_prediction(self, train_start: int, train_end: int, pred_year: int, pred_quarters: list = None):
        self.__validate_dates(start_year=train_start, end_year=pred_year)

        prices = self.__prices_data[pred_year]
        prices = prices[prices.iloc[:].notnull()]

        fig = plt.figure(num=self.__company_name + ' prediction')
        ax = plt.gca()
        fig.set_size_inches(12, 6)

        x_obs = list(range(prices.index[0], prices.index[-1] + 1))
        x_mesh, y_mean, y_cov = self.__gpr.get_eval_model(start_year=train_start, end_year=train_end,
                                                          pred_year=pred_year,
                                                          pred_quarters=pred_quarters)
        y_lower = y_mean - np.sqrt(np.diag(y_cov))
        y_upper = y_mean + np.sqrt(np.diag(y_cov))
        y_max = max(abs(min(y_lower) - 1), abs(max(y_upper) + 1))
        ax.set_ylim(bottom=-y_max, top=y_max)

        x_min, x_max = -10, self.__max_days + 10
        ax.set_xlim(left=x_min, right=x_max)

        plt.plot(x_obs, prices, color='#006699', alpha=.95, label=u'Observations ' + str(pred_year), zorder=10)
        plt.plot(x_mesh, y_mean, color='#ff0066', linestyle='--', label=u'Prediction')
        plt.fill_between(x_mesh, y_lower, y_upper,
                         alpha=.25, label='95% confidence', color='#ff0066')

        handles, labels = plt.gca().get_legend_handles_labels()
        new_labels, new_handles = [], []
        for handle, label in zip(handles, labels):
            if label not in new_labels:
                new_labels.append(label)
                new_handles.append(handle)
        plt.legend(new_handles, new_labels, bbox_to_anchor=(0.01, 0.02), loc='lower left', borderaxespad=0.)

        for i in range(0, 5):
            plt.vlines(x=self.__quarter_length * i, ymin=-y_max, ymax=y_max, color='black', linestyles='--', alpha=.6,
                       zorder=-1)
            if i < 4:
                ax.text(self.__quarter_length * i + self.__quarter_length / 2 - 5, y_max - 0.5, self.__quarters[i],
                        fontsize=12)
        plt.hlines(y=0, xmin=x_min, xmax=x_max, color='black', linestyles='--', alpha=.6, zorder=-1)

        plt.grid(True, alpha=.25)
        plt.title(self.__company_name)
        plt.xlabel('Days\n')
        plt.ylabel('Normalized price')

        plt.tight_layout()

        fname = '{}_{}_prediction.png'.format(self.__company_name, pred_year)
        fig.savefig(fname, dpi=fig.dpi)
        plt.clf()

    def show_whole_time_series(self, intermediate: bool = False):
        self.show_time_series(start_year=self.__years[0], end_year=self.__years[-1], intermediate=intermediate)

    def show_time_series(self, start_year: int, end_year: int, intermediate: bool = True):
        self.__validate_dates(start_year=start_year, end_year=end_year)

        prices_data = self.__company_handler.get_whole_prices(start_year=start_year, end_year=end_year)

        fig = plt.figure(num=self.__company_name + ' prices')
        fig.set_size_inches(12, 6)
        plt.plot(prices_data.iloc[:, 0], prices_data.iloc[:, 1], color='#006699', alpha=.95,
                 label=u'Observations ' + str(start_year) + '-' + str(end_year), zorder=10)
        ax = plt.gca()

        x_ticks = []
        for year in range(start_year, end_year + 2):
            if year == end_year + 1:
                current_date = prices_data[prices_data['Date'].dt.year == end_year].iloc[-1, 0]
            else:
                current_date = prices_data[prices_data['Date'].dt.year == year].iloc[0, 0]
            x_ticks.append(current_date)

        x_formatter = mdates.DateFormatter('%d-%m-%Y')
        ax.xaxis.set_major_formatter(x_formatter)
        if not intermediate:
            x_ticks = [x_ticks[0], x_ticks[-2], x_ticks[-1]]
            ax.set_xticks([x_ticks[0], x_ticks[-1]])
        else:
            ax.set_xticks(x_ticks)
        plt.xticks(rotation=20)
        y_min, y_max = ax.get_ylim()
        x_min, x_max = ax.get_xlim()
        ax.set_ylim(bottom=y_min, top=y_max)
        ax.set_xlim(left=x_min, right=x_max)

        for i in range(0, len(x_ticks)):
            plt.vlines(x=x_ticks[i], ymin=y_min, ymax=y_max, color='black', linestyles='--', alpha=.6,
                       zorder=-1)

        plt.grid(True, alpha=0.25)
        plt.legend()
        plt.title(self.__company_name)
        plt.ylabel('Price')

        plt.tight_layout()

        fname = '{}_{}_{}_prices.png'.format(self.__company_name, start_year, end_year)
        fig.savefig(fname, dpi=fig.dpi)
        plt.clf()

    def __validate_dates(self, start_year: int, end_year: int):
        if start_year < self.__years[0] or end_year > self.__years[-1]:
            raise ValueError('\n' +
                             'Input years out of available range! \n' +
                             'Max range available: {}-{}\n'.format(self.__years[0], self.__years[-1]) +
                             'Was: {}-{}'.format(start_year, end_year))