plt.show()
return 'draw success!'
def just_test(s, n1, n2):
data = getData(s)
label = getLabel(s)
for x in range(n1, n2):
print(label[x])
draw_by_pixel(data[x])
# print "train group:"
# just_test('train', 5, 10)
# print "test group:"
# just_test('test', 105, 110)
train_data ,train_lables = getData('train'),getLabel('train').squeeze()
test_data ,test_lables = getData('test'),getLabel('test').squeeze()
t0 =time.time()
nb_clf = GaussianNB()
nb_clf.fit(train_data,train_lables)
nb_pred = nb_clf.predict(test_data)
print("nb_clf has fitted,time cost :%.3fs"%(time.time() -t0))
print("the accuracy of Gaussian navie bayes classifier:\n",accuracy_score(test_lables, nb_pred))
from sklearn.decomposition import PCA
n_components = 100
pca = PCA(svd_solver='randomized',n_components = n_components,whiten = True).fit(train_data)
def just_test(s, n1, n2):
data = getData(s)
label = getLabel(s)
for x in range(n1, n2):
print(label[x])
draw_by_pixel(data[x])
import torch.nn as nn
from torch.autograd import Variable
import torch.nn.functional as F
import torch.utils.data as Data
from input_file import getData, getLabel
import pdb
import time
t = time.time()
INPUT_SIZE = 28 * 28
CLASS_NUM = 10
BATCH_SIZE = 100
EPOCH = 20
train_x = torch.from_numpy(getData("train") / 255.)
train_y = torch.from_numpy(getLabel("train"))
test_x = torch.from_numpy(getData("test") / 255.)
test_y = torch.from_numpy(getLabel("test"))
train_minist_dataset = Data.TensorDataset(data_tensor=train_x,
target_tensor=train_y)
test_minist_dataset = Data.TensorDataset(data_tensor=test_x,
target_tensor=test_y)
train_loader = Data.DataLoader(
dataset=train_minist_dataset,
batch_size=BATCH_SIZE,
shuffle=True,
)
test_loader = Data.DataLoader(
dataset=test_minist_dataset,