assert (j.max() == len(j) - 1).all()
j_range = len(j) - 1
self.divider = j_range / self.range
transformed = j / self.divider
transformed = transformed - self.upper
transformed = erfinv(transformed)
return transformed
is_read_data = False
if is_read_data:
from utils import read_train_test_data #,tmp_read_train_valid
X, X_test, _ = read_train_test_data()
# drop label encoding
lbl_cols = [col for col in X.columns if "_labelencod" in col]
X.drop(lbl_cols, axis=1, inplace=True)
X_test.drop(lbl_cols, axis=1, inplace=True)
# nan index
train_nan_idx = csr_matrix((np.isnan(X)).astype(int))
test_nan_idx = csr_matrix((np.isnan(X_test)).astype(int))
X = X.fillna(X.median()) #X.fillna(X.median()) # X.fillna(0)
X = X.replace(np.inf, 9999.999)
X = X.replace(-np.inf, -9999.999)
X = X.values
X_test = X_test.fillna(X_test.median()) #X_test.fillna(X_test.median())
X_test = X_test.replace(np.inf, 9999.999)
"lgbm", outputname)] = data["y_test_pred"]
to_parquet(
df_train_out,
"../stacking/oof_classification_{}_{}_{}_train.parquet".format(
"lgbm", outputname, seed))
to_parquet(
df_test_out,
"../stacking/oof_classification_{}_{}_{}_test.parquet".format(
"lgbm", outputname, seed))
try:
is_read_data = True
if is_read_data:
from utils import read_train_test_data #,tmp_read_train_valid
X_train, X_test, y_train = read_train_test_data()
if feat == "select200":
nogain_features = []
f = open("../tmp/no_gain_features_selection_model.txt")
for l in f.readlines():
nogain_features.append(l.replace("\n", ""))
f.close()
drop_cols = [
col for col in X_train.columns if col in nogain_features
]
elif feat == "nooof":
drop_cols = [col for col in X_train.columns if "oof_" in col]
else:
drop_cols = []
X_train = X_train.drop(drop_cols, axis=1)
X_test = X_test.drop(drop_cols, axis=1)
import math
from tqdm import tqdm
import os
# tf.config.experimental_run_functions_eagerly(True)
os.environ["CUDA_VISIBLE_DEVICES"] = "0"
# (train_img, train_lab),(test_img, test_lab) = utils.data_loader("CIFAR10")
model = san.san(sa_type=1, layers=(2, 1, 2, 4, 1), kernels=[3, 7, 7, 7, 7])
model.build(input_shape=(config.BATCH_SIZE, config.channels,
config.image_height, config.image_width))
model.summary()
train_img, train_lab, test_img, test_lab = utils.read_train_test_data(
"/Users/hamnamoieez/Desktop/Projects/self-attention-image-recognition/dataset"
)
train_img = utils.data_preprocess(train_img)
train_lab = utils.one_hot_encoder(train_lab)
X_train, X_val, y_train, y_val = utils.validation_data(train_img, train_lab)
train_generator, val_generator = utils.data_augmentation(
X_train, y_train, X_val, y_val)
# define loss and optimizer
loss_object = tf.keras.losses.CategoricalCrossentropy()
optimizer = tf.keras.optimizers.Adam()
train_loss = tf.keras.metrics.Mean(name='train_loss')
train_accuracy = tf.keras.metrics.CategoricalAccuracy(name='train_accuracy')
valid_loss = tf.keras.metrics.Mean(name='valid_loss')
valid_accuracy = tf.keras.metrics.CategoricalAccuracy(name='valid_accuracy')