netsatsawat

_df = np.column_stack([train_df[FT_COLS].values])
hmm_model = GaussianHMM(n_components=3, covariance_type="full", 
                        n_iter=1000, random_state=SEED).fit(_df)

hidden_states = hmm_model.predict(_df)
print("Means and vars of each hidden state")
for i in range(hmm_model.n_components):
    print(f'{i}th hidden state')
    print('mean: ', (hmm_model.means_[i]))
    print('var: ', np.diag(hmm_model.covars_[i]))

netsatsawat

states = (pd.DataFrame(hidden_states, columns=['states'], index=train_df.index)
          .join(train_df, how='inner')
          .assign(mkt_cret=train_df.sret.cumsum())
          .reset_index(drop=False)
          .rename(columns={'index':'Date'}))

sns.set(font_scale=1.5)
sns.set_style('white', style_kwds)
order = np.arange(model.n_components)
fg = sns.FacetGrid(data=states, hue='states', hue_order=order,

netsatsawat

train_df = data_df.loc[: '2019-01-01'].dropna()
test_df = data_df.loc['2019-01-01': ].dropna()
X_train = train_df[FT_COLS].values
X_test = test_df[FT_COLS].values

model = mix.GaussianMixture(n_components=N_COMPONENTS, 
                            covariance_type="full", 
                            n_init=100, 
                            random_state=SEED).fit(X_train)

netsatsawat

import numpy as np
import pandas as pd
import pandas_datareader.data as web
import scipy.stats as scs

import matplotlib as mpl
from matplotlib import cm
import matplotlib.pyplot as plt
from matplotlib.dates import YearLocator, MonthLocator
import seaborn as sns

netsatsawat

train_gen = tf.keras.preprocessing.image.ImageDataGenerator(rotation_range=40, shear_range=0.2, zoom_range=0.2, 
                                                            horizontal_flip=True, vertical_flip=True, rescale=1./255.,
                                                            validation_split=0.2)
train_generator = train_gen.flow_from_directory(TRAIN_DIR, target_size=IMG_SIZE, batch_size=32, 
                                                class_mode='categorical', subset='training')
valid_generator = train_gen.flow_from_directory(TRAIN_DIR, target_size=IMG_SIZE, batch_size=32, 
                                                class_mode='categorical', subset='validation')

cnn_model2 = tf.keras.Sequential([
    tf.keras.layers.Conv2D(32, (3, 3), activation='relu', input_shape=(150, 150, 3)), 

netsatsawat

import tensorflow as tf
import tensorflow_docs as tfdocs

cnn_model = tf.keras.Sequential([
    tf.keras.layers.Conv2D(32, (3, 3), activation='relu', input_shape=train_imgs.shape[1: 4]), 
    tf.keras.layers.MaxPooling2D(2,2),
    tf.keras.layers.Conv2D(32, (3, 3), activation='relu'),
    tf.keras.layers.MaxPooling2D(2,2),
    tf.keras.layers.Conv2D(32, (3, 3), activation='relu'),
    tf.keras.layers.MaxPooling2D(2,2),

netsatsawat

lstm_model = tf.keras.Sequential([
    layers.LSTM(32, return_sequences=True, input_shape=(n_steps, 1)),
    layers.LSTM(32, return_sequences=True),
    layers.Dropout(0.2),
    layers.LSTM(32, return_sequences=True),
    layers.LSTM(32),
    layers.Dropout(0.2),
    layers.Dense(1)
])

netsatsawat

import numpy as np
import tensorflow as tf
from tensorflow import keras
from tensorflow.keras import layers

# require for installation: !pip install -q git+https://github.com/tensorflow/docs
import tensorflow_docs as tfdocs
import tensorflow_docs.plots
import tensorflow_docs.modeling

netsatsawat

import pandas as pd
import numpy as np
import random
from sklearn.datasets import make_regression

import tensorflow as tf
from tensorflow import keras
from tensorflow.keras import layers

# require for installation: !pip install -q git+https://github.com/tensorflow/docs

netsatsawat

def _get_chunk(X, y, chunkrows):
    X_chunk, y_chunk = X[chunkrows], y[chunkrows]
    return X_chunk, y_chunk


def _iter_minibatch(X, y, chunk_size):
    '''
    Construct minibatch generator
    '''
    _start = 0