/
fit_hmm.py
32 lines (24 loc) · 1.05 KB
/
fit_hmm.py
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
import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
from hmmlearn import hmm
import argparse
import os
parser = argparse.ArgumentParser()
parser.add_argument('--dataset', default='/home/edvard/SharedWithVirtualBox/corn2015-2017/corn2013-2017.txt')
parser.add_argument('--data-subset', default='/home/edvard/SharedWithVirtualBox/corn2015-2017/corn2015-2017.txt')
parser.add_argument('--some-third-file', default='/home/edvard/SharedWithVirtualBox/corn2015-2017/corn_OHLC2013-2017.txt')
if __name__ == '__main__':
args = parser.parse_args()
data_frame = pd.read_csv(args.dataset)
subset_frame = pd.read_csv(args.data_subset)
N = 10
model = hmm.GaussianHMM(n_components=N)
#model.startprob_ = np.array([1 / N] * N)
#model.transmat_ = np.random.rand(N, N) # needs to be "ergodic" (idk what this means)
#print(data_frame.iloc[:,1])
model.fit(data_frame.iloc[:,1].to_numpy().reshape(-1, 1))
print(model.startprob_)
print(model.transmat_)
print(model.transmat_prior)
sample_x, sample_z = model.sample(100)