# -*- coding: utf-8 -*-

"""

@author: Yassine BEZZA <[email protected]>

@brief:

"""

import numpy as np

from pydd.solver import GenericSolver

from pydd.models import MLP

from pydd.connectors import ArrayConnector

from sklearn import datasets, metrics, model_selection, preprocessing

import subprocess

import sys

# Redirect stdout to a file

orig_stdout = sys.stdout

log_file = 'logs.txt'

f = open(log_file, 'w')

sys.stdout = f

# Parameters

seed = 1337

np.random.seed(seed) # for reproducibility

n_classes = 10

params = {"port": 8080, "nclasses": n_classes, "gpu": True}

split_params = {"test_size": 0.2, "random_state": seed}

# create dataset

X, y = datasets.load_digits(n_class=n_classes, return_X_y=True)

X = preprocessing.StandardScaler().fit_transform(X)

xtr, xte, ytr, yte = model_selection.train_test_split(X, y, **split_params)

# Define models and class weights

clf = MLP(**params)

solver = GenericSolver(iterations=5000, solver_type="SGD", base_lr=0.01, gamma=0.1, stepsize=30, momentum=0.9)

# one class weight value for each class

class_weights = [1., 1., 1., 1., 1., 1., 1., 1., 1., 1]

train_data, test_data = ArrayConnector(xtr, ytr), ArrayConnector(xte, yte)

# Start visdom on another process

#start_visdom = ["python'", "-m", "visdom.server"]

#subprocess.Popen(["python", "-m", "visdom.server"])

# Listen visdom on logs file

#listen_logs = ["pydd", "--log_dir", log_file]

#subprocess.Popen(listen_logs)

# Start listening to the port

logs = clf.fit(train_data, validation_data=[test_data], solver=solver, class_weights=class_weights, batch_size=128)

yte_pred = clf.predict(test_data)

report = metrics.classification_report(yte, yte_pred)

print(report)

# Close the output file

sys.stdout = orig_stdout

f.close()