diff --git a/modules/deeplearning/srcnn_l1b_l2_v2.py b/modules/deeplearning/srcnn_l1b_l2_v2.py
new file mode 100644
index 0000000000000000000000000000000000000000..a298ff87a4112f4247c917973fbc4fdd3fc3e999
--- /dev/null
+++ b/modules/deeplearning/srcnn_l1b_l2_v2.py
@@ -0,0 +1,904 @@
+import glob
+import tensorflow as tf
+
+import util.util
+from util.setup import logdir, modeldir, cachepath, now, ancillary_path
+from util.util import EarlyStop, normalize, denormalize, resample, resample_2d_linear, resample_one,\
+ resample_2d_linear_one, get_grid_values_all, add_noise, smooth_2d, smooth_2d_single, median_filter_2d,\
+ median_filter_2d_single, downscale_2x
+import os, datetime
+import numpy as np
+import pickle
+import h5py
+from scipy.ndimage import gaussian_filter
+
+# L1B M/I-bands: /apollo/cloud/scratch/cwhite/VIIRS_HRES/2019/2019_01_01/
+# CLAVRx: /apollo/cloud/scratch/Satellite_Output/VIIRS_HRES/2019/2019_01_01/
+# /apollo/cloud/scratch/Satellite_Output/andi/NEW/VIIRS_HRES/2019
+
+LOG_DEVICE_PLACEMENT = False
+
+PROC_BATCH_SIZE = 4
+PROC_BATCH_BUFFER_SIZE = 50000
+
+NumClasses = 2
+if NumClasses == 2:
+ NumLogits = 1
+else:
+ NumLogits = NumClasses
+
+BATCH_SIZE = 128
+NUM_EPOCHS = 80
+
+TRACK_MOVING_AVERAGE = False
+EARLY_STOP = True
+
+NOISE_TRAINING = False
+NOISE_STDDEV = 0.01
+DO_AUGMENT = True
+
+DO_SMOOTH = False
+SIGMA = 1.0
+DO_ZERO_OUT = False
+DO_ESPCN = False # Note: If True, cannot do mixed resolution input fields (Adjust accordingly below)
+
+# setup scaling parameters dictionary
+mean_std_dct = {}
+mean_std_file = ancillary_path+'mean_std_lo_hi_l2.pkl'
+f = open(mean_std_file, 'rb')
+mean_std_dct_l2 = pickle.load(f)
+f.close()
+
+mean_std_file = ancillary_path+'mean_std_lo_hi_l1b.pkl'
+f = open(mean_std_file, 'rb')
+mean_std_dct_l1b = pickle.load(f)
+f.close()
+
+mean_std_dct.update(mean_std_dct_l1b)
+mean_std_dct.update(mean_std_dct_l2)
+
+IMG_DEPTH = 1
+# label_param = 'cloud_fraction'
+label_param = 'cld_opd_dcomp'
+# label_param = 'cloud_probability'
+
+params = ['temp_11_0um_nom', 'temp_12_0um_nom', 'refl_0_65um_nom', label_param]
+data_params_half = ['temp_11_0um_nom']
+data_params_full = ['refl_0_65um_nom']
+
+label_idx = params.index(label_param)
+
+print('data_params_half: ', data_params_half)
+print('data_params_full: ', data_params_full)
+print('label_param: ', label_param)
+
+KERNEL_SIZE = 3 # target size: (128, 128)
+N = 1
+
+if KERNEL_SIZE == 3:
+ slc_x = slice(1, N*64 + 4)
+ slc_y = slice(1, N*64 + 4)
+ x_2 = np.arange(N*64 + 3)
+ y_2 = np.arange(N*64 + 3)
+ t = np.arange(0.5, N*64 + 2.5, 0.5)
+ s = np.arange(0.5, N*64 + 2.5, 0.5)
+ x_k = slice(1, N*128 + 3)
+ y_k = slice(1, N*128 + 3)
+ x_128 = slice(4, N*128 + 4)
+ y_128 = slice(4, N*128 + 4)
+elif KERNEL_SIZE == 5:
+ # slc_x = slice(3, 135)
+ # slc_y = slice(3, 135)
+ # slc_x_2 = slice(2, 137, 2)
+ # slc_y_2 = slice(2, 137, 2)
+ # x_128 = slice(5, 133)
+ # y_128 = slice(5, 133)
+ # t = np.arange(1, 67, 0.5)
+ # s = np.arange(1, 67, 0.5)
+ # x_2 = np.arange(68)
+ # y_2 = np.arange(68)
+ pass # Not yet
+# ----------------------------------------
+# Exp for ESPCN version
+if DO_ESPCN:
+ slc_x_2 = slice(0, 132, 2)
+ slc_y_2 = slice(0, 132, 2)
+ x_128 = slice(2, 130)
+ y_128 = slice(2, 130)
+
+
+def build_residual_conv2d_block(conv, num_filters, block_name, activation=tf.nn.relu, padding='SAME',
+ kernel_initializer='he_uniform', scale=None, kernel_size=3,
+ do_drop_out=True, drop_rate=0.5, do_batch_norm=True):
+
+ with tf.name_scope(block_name):
+ skip = tf.keras.layers.Conv2D(num_filters, kernel_size=kernel_size, padding=padding, kernel_initializer=kernel_initializer, activation=activation)(conv)
+ skip = tf.keras.layers.Conv2D(num_filters, kernel_size=kernel_size, padding=padding, activation=None)(skip)
+
+ if scale is not None:
+ skip = tf.keras.layers.Lambda(lambda x: x * scale)(skip)
+
+ if do_drop_out:
+ skip = tf.keras.layers.Dropout(drop_rate)(skip)
+
+ if do_batch_norm:
+ skip = tf.keras.layers.BatchNormalization()(skip)
+
+ conv = conv + skip
+ print(block_name+':', conv.shape)
+
+ return conv
+
+
+def upsample(tmp):
+ # tmp = tmp[:, slc_y_2, slc_x_2]
+ tmp = tmp[:, slc_y, slc_x]
+ tmp = resample_2d_linear(x_2, y_2, tmp, t, s)
+ tmp = tmp[:, y_k, x_k]
+ return tmp
+
+
+def upsample_nearest(tmp):
+ bsize = tmp.shape[0]
+ tmp_2 = tmp[:, slc_y_2, slc_x_2]
+ up = np.zeros(bsize, t.size, s.size)
+ for k in range(bsize):
+ for j in range(t.size/2):
+ for i in range(s.size/2):
+ up[k, j, i] = tmp_2[k, j, i]
+ up[k, j, i+1] = tmp_2[k, j, i]
+ up[k, j+1, i] = tmp_2[k, j, i]
+ up[k, j+1, i+1] = tmp_2[k, j, i]
+ return up
+
+
+class SRCNN:
+
+ def __init__(self):
+
+ self.train_data = None
+ self.train_label = None
+ self.test_data = None
+ self.test_label = None
+ self.test_data_denorm = None
+
+ self.train_dataset = None
+ self.inner_train_dataset = None
+ self.test_dataset = None
+ self.eval_dataset = None
+ self.X_img = None
+ self.X_prof = None
+ self.X_u = None
+ self.X_v = None
+ self.X_sfc = None
+ self.inputs = []
+ self.y = None
+ self.handle = None
+ self.inner_handle = None
+ self.in_mem_batch = None
+
+ self.h5f_l1b_trn = None
+ self.h5f_l1b_tst = None
+ self.h5f_l2_trn = None
+ self.h5f_l2_tst = None
+
+ self.logits = None
+
+ self.predict_data = None
+ self.predict_dataset = None
+ self.mean_list = None
+ self.std_list = None
+
+ self.training_op = None
+ self.correct = None
+ self.accuracy = None
+ self.loss = None
+ self.pred_class = None
+ self.variable_averages = None
+
+ self.global_step = None
+
+ self.writer_train = None
+ self.writer_valid = None
+ self.writer_train_valid_loss = None
+
+ self.OUT_OF_RANGE = False
+
+ self.abi = None
+ self.temp = None
+ self.wv = None
+ self.lbfp = None
+ self.sfc = None
+
+ self.in_mem_data_cache = {}
+ self.in_mem_data_cache_test = {}
+
+ self.model = None
+ self.optimizer = None
+ self.ema = None
+ self.train_loss = None
+ self.train_accuracy = None
+ self.test_loss = None
+ self.test_accuracy = None
+ self.test_auc = None
+ self.test_recall = None
+ self.test_precision = None
+ self.test_confusion_matrix = None
+ self.test_true_pos = None
+ self.test_true_neg = None
+ self.test_false_pos = None
+ self.test_false_neg = None
+
+ self.test_labels = []
+ self.test_preds = []
+ self.test_probs = None
+
+ self.learningRateSchedule = None
+ self.num_data_samples = None
+ self.initial_learning_rate = None
+
+ self.data_dct = None
+ self.train_data_files = None
+ self.train_label_files = None
+ self.test_data_files = None
+ self.test_label_files = None
+
+ self.train_data_nda = None
+ self.train_label_nda = None
+ self.test_data_nda = None
+ self.test_label_nda = None
+
+ self.n_chans = len(data_params_half) + len(data_params_full) + 1
+
+ self.X_img = tf.keras.Input(shape=(None, None, self.n_chans))
+
+ self.inputs.append(self.X_img)
+
+ tf.debugging.set_log_device_placement(LOG_DEVICE_PLACEMENT)
+
+ def get_in_mem_data_batch(self, idxs, is_training):
+ if is_training:
+ files = self.train_data_files
+ else:
+ files = self.test_data_files
+
+ data_s = []
+ for k in idxs:
+ f = files[k]
+ try:
+ nda = np.load(f)
+ except Exception:
+ print(f)
+ continue
+ data_s.append(nda)
+ input_data = np.concatenate(data_s)
+
+ DO_ADD_NOISE = False
+ if is_training and NOISE_TRAINING:
+ DO_ADD_NOISE = True
+
+ data_norm = []
+ for param in data_params_half:
+ idx = params.index(param)
+ tmp = input_data[:, idx, :, :]
+ tmp = tmp.copy()
+ tmp = np.where(np.isnan(tmp), 0, tmp)
+ if DO_ESPCN:
+ tmp = tmp[:, slc_y_2, slc_x_2]
+ else: # Half res upsampled to full res:
+ tmp = upsample(tmp)
+ tmp = normalize(tmp, param, mean_std_dct)
+ if DO_ADD_NOISE:
+ tmp = add_noise(tmp, noise_scale=NOISE_STDDEV)
+ data_norm.append(tmp)
+
+ for param in data_params_full:
+ idx = params.index(param)
+ tmp = input_data[:, idx, :, :]
+ tmp = tmp.copy()
+ tmp = np.where(np.isnan(tmp), 0, tmp)
+ # Full res:
+ tmp = tmp[:, slc_y, slc_x]
+ tmp = normalize(tmp, param, mean_std_dct)
+ if DO_ADD_NOISE:
+ tmp = add_noise(tmp, noise_scale=NOISE_STDDEV)
+ data_norm.append(tmp)
+ # ---------------------------------------------------
+ tmp = input_data[:, label_idx, :, :]
+ tmp = tmp.copy()
+ tmp = np.where(np.isnan(tmp), 0, tmp)
+ if DO_SMOOTH:
+ tmp = smooth_2d(tmp, sigma=SIGMA)
+ # tmp = median_filter_2d(tmp)
+ if DO_ESPCN:
+ tmp = tmp[:, slc_y_2, slc_x_2]
+ else: # Half res upsampled to full res:
+ tmp = upsample(tmp)
+ if label_param != 'cloud_probability':
+ tmp = normalize(tmp, label_param, mean_std_dct)
+ if DO_ADD_NOISE:
+ tmp = add_noise(tmp, noise_scale=NOISE_STDDEV)
+ else:
+ if DO_ADD_NOISE:
+ tmp = add_noise(tmp, noise_scale=NOISE_STDDEV)
+ tmp = np.where(tmp < 0.0, 0.0, tmp)
+ tmp = np.where(tmp > 1.0, 1.0, tmp)
+ data_norm.append(tmp)
+ # ---------
+ data = np.stack(data_norm, axis=3)
+ data = data.astype(np.float32)
+ # -----------------------------------------------------
+ # -----------------------------------------------------
+ label = input_data[:, label_idx, :, :]
+ label = label.copy()
+ if DO_SMOOTH:
+ label = np.where(np.isnan(label), 0, label)
+ label = smooth_2d(label, sigma=SIGMA)
+ # label = median_filter_2d(label)
+ label = label[:, y_128, x_128]
+
+ if label_param != 'cloud_probability':
+ label = normalize(label, label_param, mean_std_dct)
+ else:
+ label = np.where(np.isnan(label), 0, label)
+ label = np.expand_dims(label, axis=3)
+
+ data = data.astype(np.float32)
+ label = label.astype(np.float32)
+
+ if is_training and DO_AUGMENT:
+ data_ud = np.flip(data, axis=1)
+ label_ud = np.flip(label, axis=1)
+
+ data_lr = np.flip(data, axis=2)
+ label_lr = np.flip(label, axis=2)
+
+ data = np.concatenate([data, data_ud, data_lr])
+ label = np.concatenate([label, label_ud, label_lr])
+
+ return data, label
+
+ def get_in_mem_data_batch_train(self, idxs):
+ return self.get_in_mem_data_batch(idxs, True)
+
+ def get_in_mem_data_batch_test(self, idxs):
+ return self.get_in_mem_data_batch(idxs, False)
+
+ @tf.function(input_signature=[tf.TensorSpec(None, tf.int32)])
+ def data_function(self, indexes):
+ out = tf.numpy_function(self.get_in_mem_data_batch_train, [indexes], [tf.float32, tf.float32])
+ return out
+
+ @tf.function(input_signature=[tf.TensorSpec(None, tf.int32)])
+ def data_function_test(self, indexes):
+ out = tf.numpy_function(self.get_in_mem_data_batch_test, [indexes], [tf.float32, tf.float32])
+ return out
+
+ def get_train_dataset(self, indexes):
+ indexes = list(indexes)
+
+ dataset = tf.data.Dataset.from_tensor_slices(indexes)
+ dataset = dataset.batch(PROC_BATCH_SIZE)
+ dataset = dataset.map(self.data_function, num_parallel_calls=8)
+ dataset = dataset.cache()
+ if DO_AUGMENT:
+ dataset = dataset.shuffle(PROC_BATCH_BUFFER_SIZE)
+ dataset = dataset.prefetch(buffer_size=1)
+ self.train_dataset = dataset
+
+ def get_test_dataset(self, indexes):
+ indexes = list(indexes)
+
+ dataset = tf.data.Dataset.from_tensor_slices(indexes)
+ dataset = dataset.batch(PROC_BATCH_SIZE)
+ dataset = dataset.map(self.data_function_test, num_parallel_calls=8)
+ dataset = dataset.cache()
+ self.test_dataset = dataset
+
+ def setup_pipeline(self, train_data_files, test_data_files, num_train_samples):
+
+ self.train_data_files = train_data_files
+ self.test_data_files = test_data_files
+
+ trn_idxs = np.arange(len(train_data_files))
+ np.random.shuffle(trn_idxs)
+ tst_idxs = np.arange(len(test_data_files))
+
+ self.get_train_dataset(trn_idxs)
+ self.get_test_dataset(tst_idxs)
+
+ self.num_data_samples = num_train_samples # approximately
+
+ print('datetime: ', now)
+ print('training and test data: ')
+ print('---------------------------')
+ print('num train samples: ', self.num_data_samples)
+ print('BATCH SIZE: ', BATCH_SIZE)
+ print('num test samples: ', tst_idxs.shape[0])
+ print('setup_pipeline: Done')
+
+ def setup_test_pipeline(self, test_data_files):
+ self.test_data_files = test_data_files
+ tst_idxs = np.arange(len(test_data_files))
+ self.get_test_dataset(tst_idxs)
+ print('setup_test_pipeline: Done')
+
+ def build_srcnn(self, do_drop_out=False, do_batch_norm=False, drop_rate=0.5, factor=2):
+ print('build_cnn')
+ padding = "SAME"
+
+ # activation = tf.nn.relu
+ # activation = tf.nn.elu
+ activation = tf.nn.relu
+ momentum = 0.99
+
+ num_filters = 64
+
+ input_2d = self.inputs[0]
+ print('input: ', input_2d.shape)
+
+ conv = conv_b = tf.keras.layers.Conv2D(num_filters, kernel_size=KERNEL_SIZE, kernel_initializer='he_uniform', activation=activation, padding='VALID')(input_2d)
+ print(conv.shape)
+
+ # if NOISE_TRAINING:
+ # conv = conv_b = tf.keras.layers.GaussianNoise(stddev=NOISE_STDDEV)(conv)
+
+ scale = 0.2
+
+ conv_b = build_residual_conv2d_block(conv_b, num_filters, 'Residual_Block_1', kernel_size=KERNEL_SIZE, scale=scale)
+
+ conv_b = build_residual_conv2d_block(conv_b, num_filters, 'Residual_Block_2', kernel_size=KERNEL_SIZE, scale=scale)
+
+ conv_b = build_residual_conv2d_block(conv_b, num_filters, 'Residual_Block_3', kernel_size=KERNEL_SIZE, scale=scale)
+
+ conv_b = build_residual_conv2d_block(conv_b, num_filters, 'Residual_Block_4', kernel_size=KERNEL_SIZE, scale=scale)
+
+ conv_b = build_residual_conv2d_block(conv_b, num_filters, 'Residual_Block_5', kernel_size=KERNEL_SIZE, scale=scale)
+
+ conv_b = build_residual_conv2d_block(conv_b, num_filters, 'Residual_Block_6', kernel_size=KERNEL_SIZE, scale=scale)
+
+ conv_b = tf.keras.layers.Conv2D(num_filters, kernel_size=3, strides=1, activation=activation, kernel_initializer='he_uniform', padding=padding)(conv_b)
+
+ conv = conv + conv_b
+ print(conv.shape)
+
+ if not DO_ESPCN:
+ # This is effectively a Dense layer
+ self.logits = tf.keras.layers.Conv2D(1, kernel_size=1, strides=1, padding=padding, name='regression')(conv)
+ else:
+ conv = tf.keras.layers.Conv2D(num_filters * (factor ** 2), 3, padding=padding, activation=activation)(conv)
+ print(conv.shape)
+ conv = tf.nn.depth_to_space(conv, factor)
+ print(conv.shape)
+ self.logits = tf.keras.layers.Conv2D(IMG_DEPTH, kernel_size=3, strides=1, padding=padding, name='regression')(conv)
+ print(self.logits.shape)
+
+ def build_training(self):
+ # if NumClasses == 2:
+ # self.loss = tf.keras.losses.BinaryCrossentropy(from_logits=False) # for two-class only
+ # else:
+ # self.loss = tf.keras.losses.SparseCategoricalCrossentropy(from_logits=False) # For multi-class
+ # self.loss = tf.keras.losses.MeanAbsoluteError() # Regression
+ self.loss = tf.keras.losses.MeanSquaredError() # Regression
+
+ # decayed_learning_rate = learning_rate * decay_rate ^ (global_step / decay_steps)
+ initial_learning_rate = 0.005
+ decay_rate = 0.95
+ steps_per_epoch = int(self.num_data_samples/BATCH_SIZE) # one epoch
+ decay_steps = int(steps_per_epoch)
+ print('initial rate, decay rate, steps/epoch, decay steps: ', initial_learning_rate, decay_rate, steps_per_epoch, decay_steps)
+
+ self.learningRateSchedule = tf.keras.optimizers.schedules.ExponentialDecay(initial_learning_rate, decay_steps, decay_rate)
+
+ optimizer = tf.keras.optimizers.Adam(learning_rate=self.learningRateSchedule)
+
+ if TRACK_MOVING_AVERAGE:
+ # Not really sure this works properly (from tfa)
+ # optimizer = tfa.optimizers.MovingAverage(optimizer)
+ self.ema = tf.train.ExponentialMovingAverage(decay=0.9999)
+
+ self.optimizer = optimizer
+ self.initial_learning_rate = initial_learning_rate
+
+ def build_evaluation(self):
+ self.train_accuracy = tf.keras.metrics.MeanAbsoluteError(name='train_accuracy')
+ self.test_accuracy = tf.keras.metrics.MeanAbsoluteError(name='test_accuracy')
+ self.train_loss = tf.keras.metrics.Mean(name='train_loss')
+ self.test_loss = tf.keras.metrics.Mean(name='test_loss')
+
+ @tf.function
+ def train_step(self, mini_batch):
+ inputs = [mini_batch[0]]
+ labels = mini_batch[1]
+ with tf.GradientTape() as tape:
+ pred = self.model(inputs, training=True)
+ loss = self.loss(labels, pred)
+ total_loss = loss
+ if len(self.model.losses) > 0:
+ reg_loss = tf.math.add_n(self.model.losses)
+ total_loss = loss + reg_loss
+ gradients = tape.gradient(total_loss, self.model.trainable_variables)
+ self.optimizer.apply_gradients(zip(gradients, self.model.trainable_variables))
+ if TRACK_MOVING_AVERAGE:
+ self.ema.apply(self.model.trainable_variables)
+
+ self.train_loss(loss)
+ self.train_accuracy(labels, pred)
+
+ return loss
+
+ @tf.function
+ def test_step(self, mini_batch):
+ inputs = [mini_batch[0]]
+ labels = mini_batch[1]
+ pred = self.model(inputs, training=False)
+ t_loss = self.loss(labels, pred)
+
+ self.test_loss(t_loss)
+ self.test_accuracy(labels, pred)
+
+ def predict(self, mini_batch):
+ inputs = [mini_batch[0]]
+ labels = mini_batch[1]
+ pred = self.model(inputs, training=False)
+ t_loss = self.loss(labels, pred)
+
+ self.test_labels.append(labels)
+ self.test_preds.append(pred.numpy())
+
+ self.test_loss(t_loss)
+ self.test_accuracy(labels, pred)
+
+ def reset_test_metrics(self):
+ self.test_loss.reset_states()
+ self.test_accuracy.reset_states()
+
+ def get_metrics(self):
+ recall = self.test_recall.result()
+ precsn = self.test_precision.result()
+ f1 = 2 * (precsn * recall) / (precsn + recall)
+
+ tn = self.test_true_neg.result()
+ tp = self.test_true_pos.result()
+ fn = self.test_false_neg.result()
+ fp = self.test_false_pos.result()
+
+ mcc = ((tp * tn) - (fp * fn)) / np.sqrt((tp + fp) * (tp + fn) * (tn + fp) * (tn + fn))
+ return f1, mcc
+
+ def do_training(self, ckpt_dir=None):
+
+ if ckpt_dir is None:
+ if not os.path.exists(modeldir):
+ os.mkdir(modeldir)
+ ckpt = tf.train.Checkpoint(step=tf.Variable(1), model=self.model)
+ ckpt_manager = tf.train.CheckpointManager(ckpt, modeldir, max_to_keep=3)
+ else:
+ ckpt = tf.train.Checkpoint(step=tf.Variable(1), model=self.model)
+ ckpt_manager = tf.train.CheckpointManager(ckpt, ckpt_dir, max_to_keep=3)
+ ckpt.restore(ckpt_manager.latest_checkpoint)
+
+ self.writer_train = tf.summary.create_file_writer(os.path.join(logdir, 'plot_train'))
+ self.writer_valid = tf.summary.create_file_writer(os.path.join(logdir, 'plot_valid'))
+ self.writer_train_valid_loss = tf.summary.create_file_writer(os.path.join(logdir, 'plot_train_valid_loss'))
+
+ step = 0
+ total_time = 0
+ best_test_loss = np.finfo(dtype=np.float).max
+
+ if EARLY_STOP:
+ es = EarlyStop()
+
+ for epoch in range(NUM_EPOCHS):
+ self.train_loss.reset_states()
+ self.train_accuracy.reset_states()
+
+ t0 = datetime.datetime.now().timestamp()
+
+ proc_batch_cnt = 0
+ n_samples = 0
+
+ for data, label in self.train_dataset:
+ trn_ds = tf.data.Dataset.from_tensor_slices((data, label))
+ trn_ds = trn_ds.batch(BATCH_SIZE)
+ for mini_batch in trn_ds:
+ if self.learningRateSchedule is not None:
+ loss = self.train_step(mini_batch)
+
+ if (step % 100) == 0:
+
+ with self.writer_train.as_default():
+ tf.summary.scalar('loss_trn', loss.numpy(), step=step)
+ tf.summary.scalar('learning_rate', self.optimizer._decayed_lr('float32').numpy(), step=step)
+ tf.summary.scalar('num_train_steps', step, step=step)
+ tf.summary.scalar('num_epochs', epoch, step=step)
+
+ self.reset_test_metrics()
+ for data_tst, label_tst in self.test_dataset:
+ tst_ds = tf.data.Dataset.from_tensor_slices((data_tst, label_tst))
+ tst_ds = tst_ds.batch(BATCH_SIZE)
+ for mini_batch_test in tst_ds:
+ self.test_step(mini_batch_test)
+
+ with self.writer_valid.as_default():
+ tf.summary.scalar('loss_val', self.test_loss.result(), step=step)
+ tf.summary.scalar('acc_val', self.test_accuracy.result(), step=step)
+
+ with self.writer_train_valid_loss.as_default():
+ tf.summary.scalar('loss_trn', loss.numpy(), step=step)
+ tf.summary.scalar('loss_val', self.test_loss.result(), step=step)
+
+ print('****** test loss, acc, lr: ', self.test_loss.result().numpy(), self.test_accuracy.result().numpy(),
+ self.optimizer._decayed_lr('float32').numpy())
+
+ step += 1
+ print('train loss: ', loss.numpy())
+
+ proc_batch_cnt += 1
+ n_samples += data.shape[0]
+ print('proc_batch_cnt: ', proc_batch_cnt, n_samples)
+
+ t1 = datetime.datetime.now().timestamp()
+ print('End of Epoch: ', epoch+1, 'elapsed time: ', (t1-t0))
+ total_time += (t1-t0)
+
+ self.reset_test_metrics()
+ for data, label in self.test_dataset:
+ ds = tf.data.Dataset.from_tensor_slices((data, label))
+ ds = ds.batch(BATCH_SIZE)
+ for mini_batch in ds:
+ self.test_step(mini_batch)
+
+ print('loss, acc: ', self.test_loss.result().numpy(), self.test_accuracy.result().numpy())
+ print('------------------------------------------------------')
+
+ tst_loss = self.test_loss.result().numpy()
+ if tst_loss < best_test_loss:
+ best_test_loss = tst_loss
+ ckpt_manager.save()
+
+ if EARLY_STOP and es.check_stop(tst_loss):
+ break
+
+ print('total time: ', total_time)
+ self.writer_train.close()
+ self.writer_valid.close()
+ self.writer_train_valid_loss.close()
+
+ # f = open(home_dir+'/best_stats_'+now+'.pkl', 'wb')
+ # pickle.dump((best_test_loss, best_test_acc, best_test_recall, best_test_precision, best_test_auc, best_test_f1, best_test_mcc), f)
+ # f.close()
+
+ def build_model(self):
+ self.build_srcnn()
+ self.model = tf.keras.Model(self.inputs, self.logits)
+
+ def restore(self, ckpt_dir):
+
+ ckpt = tf.train.Checkpoint(step=tf.Variable(1), model=self.model)
+ ckpt_manager = tf.train.CheckpointManager(ckpt, ckpt_dir, max_to_keep=3)
+ ckpt.restore(ckpt_manager.latest_checkpoint)
+
+ self.reset_test_metrics()
+
+ for data, label in self.test_dataset:
+ ds = tf.data.Dataset.from_tensor_slices((data, label))
+ ds = ds.batch(BATCH_SIZE)
+ for mini_batch_test in ds:
+ self.predict(mini_batch_test)
+
+ print('loss, acc: ', self.test_loss.result().numpy(), self.test_accuracy.result().numpy())
+
+ labels = np.concatenate(self.test_labels)
+ preds = np.concatenate(self.test_preds)
+ print(labels.shape, preds.shape)
+
+ labels_denorm = denormalize(labels, label_param, mean_std_dct)
+ preds_denorm = denormalize(preds, label_param, mean_std_dct)
+
+ return labels_denorm, preds_denorm
+
+ def do_evaluate(self, data, ckpt_dir):
+
+ ckpt = tf.train.Checkpoint(step=tf.Variable(1), model=self.model)
+ ckpt_manager = tf.train.CheckpointManager(ckpt, ckpt_dir, max_to_keep=3)
+ ckpt.restore(ckpt_manager.latest_checkpoint)
+
+ self.reset_test_metrics()
+
+ pred = self.model([data], training=False)
+ self.test_probs = pred
+ pred = pred.numpy()
+ if label_param != 'cloud_probability':
+ pred = denormalize(pred, label_param, mean_std_dct)
+
+ return pred
+
+ def run(self, directory, ckpt_dir=None, num_data_samples=50000):
+ train_data_files = glob.glob(directory+'data_train_*.npy')
+ valid_data_files = glob.glob(directory+'data_valid_*.npy')
+
+ self.setup_pipeline(train_data_files, valid_data_files, num_data_samples)
+ self.build_model()
+ self.build_training()
+ self.build_evaluation()
+ self.do_training(ckpt_dir=ckpt_dir)
+
+ def run_restore(self, directory, ckpt_dir):
+ valid_data_files = glob.glob(directory + 'data_valid*.npy')
+ self.num_data_samples = 1000
+ self.setup_test_pipeline(valid_data_files)
+ self.build_model()
+ self.build_training()
+ self.build_evaluation()
+ return self.restore(ckpt_dir)
+
+ def run_evaluate(self, data, ckpt_dir):
+ data = tf.convert_to_tensor(data, dtype=tf.float32)
+ self.num_data_samples = 80000
+ self.build_model()
+ self.build_training()
+ self.build_evaluation()
+ return self.do_evaluate(data, ckpt_dir)
+
+
+def run_restore_static(directory, ckpt_dir, out_file=None):
+ nn = SRCNN()
+ labels_denorm, preds_denorm = nn.run_restore(directory, ckpt_dir)
+ if out_file is not None:
+ np.save(out_file, [labels_denorm, preds_denorm])
+
+
+def run_evaluate_static(in_file, out_file, ckpt_dir):
+ N = 10
+
+ slc_x = slice(2, N*128 + 4)
+ slc_y = slice(2, N*128 + 4)
+ slc_x_2 = slice(1, N*128 + 6, 2)
+ slc_y_2 = slice(1, N*128 + 6, 2)
+ x_2 = np.arange(int((N*128)/2) + 3)
+ y_2 = np.arange(int((N*128)/2) + 3)
+ t = np.arange(0, int((N*128)/2) + 3, 0.5)
+ s = np.arange(0, int((N*128)/2) + 3, 0.5)
+ x_k = slice(1, N*128 + 3)
+ y_k = slice(1, N*128 + 3)
+ x_128 = slice(3, N*128 + 3)
+ y_128 = slice(3, N*128 + 3)
+
+ sub_y, sub_x = (N * 128) + 10, (N * 128) + 10
+ y_0, x_0, = 3232 - int(sub_y/2), 3200 - int(sub_x/2)
+
+ h5f = h5py.File(in_file, 'r')
+
+ grd_a = get_grid_values_all(h5f, 'temp_11_0um_nom')
+ grd_a = grd_a[y_0:y_0+sub_y, x_0:x_0+sub_x]
+ grd_a = grd_a.copy()
+ grd_a = np.where(np.isnan(grd_a), 0, grd_a)
+ hr_grd_a = grd_a.copy()
+ hr_grd_a = hr_grd_a[y_128, x_128]
+ # Full res:
+ # grd_a = grd_a[slc_y, slc_x]
+ # Half res:
+ grd_a = grd_a[slc_y_2, slc_x_2]
+ grd_a = resample_2d_linear_one(x_2, y_2, grd_a, t, s)
+ grd_a = grd_a[y_k, x_k]
+ grd_a = normalize(grd_a, 'temp_11_0um_nom', mean_std_dct)
+ # ------------------------------------------------------
+ grd_b = get_grid_values_all(h5f, 'refl_0_65um_nom')
+ grd_b = grd_b[y_0:y_0+sub_y, x_0:x_0+sub_x]
+ grd_b = grd_b.copy()
+ grd_b = np.where(np.isnan(grd_b), 0, grd_b)
+ hr_grd_b = grd_b.copy()
+ hr_grd_b = hr_grd_b[y_128, x_128]
+ grd_b = grd_b[slc_y, slc_x]
+ grd_b = normalize(grd_b, 'refl_0_65um_nom', mean_std_dct)
+
+ grd_c = get_grid_values_all(h5f, label_param)
+ grd_c = grd_c[y_0:y_0+sub_y, x_0:x_0+sub_x]
+ hr_grd_c = grd_c.copy()
+ hr_grd_c = np.where(np.isnan(hr_grd_c), 0, grd_c)
+ hr_grd_c = hr_grd_c[y_128, x_128]
+ # hr_grd_c = smooth_2d_single(hr_grd_c, sigma=1.0)
+ grd_c = np.where(np.isnan(grd_c), 0, grd_c)
+ grd_c = grd_c.copy()
+ # grd_c = smooth_2d_single(grd_c, sigma=1.0)
+ grd_c = grd_c[slc_y_2, slc_x_2]
+ grd_c = resample_2d_linear_one(x_2, y_2, grd_c, t, s)
+ grd_c = grd_c[y_k, x_k]
+ if label_param != 'cloud_probability':
+ grd_c = normalize(grd_c, label_param, mean_std_dct)
+
+ data = np.stack([grd_a, grd_b, grd_c], axis=2)
+ data = np.expand_dims(data, axis=0)
+
+ h5f.close()
+
+ nn = SRCNN()
+ out_sr = nn.run_evaluate(data, ckpt_dir)
+ if out_file is not None:
+ np.save(out_file, (out_sr[0, :, :, 0], hr_grd_a, hr_grd_b, hr_grd_c))
+ else:
+ return out_sr, hr_grd_a, hr_grd_b, hr_grd_c
+
+
+def analyze(file='/Users/tomrink/cld_opd_out.npy'):
+ # Save this:
+ # nn.test_data_files = glob.glob('/Users/tomrink/data/clavrx_opd_valid_DAY/data_valid*.npy')
+ # idxs = np.arange(50)
+ # dat, lbl = nn.get_in_mem_data_batch(idxs, False)
+ # tmp = dat[:, 1:128, 1:128, 1]
+ # tmp = dat[:, 1:129, 1:129, 1]
+
+ tup = np.load(file, allow_pickle=True)
+ lbls = tup[0]
+ pred = tup[1]
+
+ lbls = lbls[:, :, :, 0]
+ pred = pred[:, :, :, 0]
+ print('Total num pixels: ', lbls.size)
+
+ pred = pred.flatten()
+ pred = np.where(pred < 0.0, 0.0, pred)
+ lbls = lbls.flatten()
+ diff = pred - lbls
+
+ mae = (np.sum(np.abs(diff))) / diff.size
+ print('MAE: ', mae)
+
+ bin_edges = []
+ bin_ranges = []
+
+ bin_ranges.append([0.0, 5.0])
+ bin_edges.append(0.0)
+
+ bin_ranges.append([5.0, 10.0])
+ bin_edges.append(5.0)
+
+ bin_ranges.append([10.0, 15.0])
+ bin_edges.append(10.0)
+
+ bin_ranges.append([15.0, 20.0])
+ bin_edges.append(15.0)
+
+ bin_ranges.append([20.0, 30.0])
+ bin_edges.append(20.0)
+
+ bin_ranges.append([30.0, 40.0])
+ bin_edges.append(30.0)
+
+ bin_ranges.append([40.0, 60.0])
+ bin_edges.append(40.0)
+
+ bin_ranges.append([60.0, 80.0])
+ bin_edges.append(60.0)
+
+ bin_ranges.append([80.0, 100.0])
+ bin_edges.append(80.0)
+
+ bin_ranges.append([100.0, 120.0])
+ bin_edges.append(100.0)
+
+ bin_ranges.append([120.0, 140.0])
+ bin_edges.append(120.0)
+
+ bin_ranges.append([140.0, 160.0])
+ bin_edges.append(140.0)
+
+ bin_edges.append(160.0)
+
+ diff_by_value_bins = util.util.bin_data_by(diff, lbls, bin_ranges)
+
+ values = []
+ for k in range(len(bin_ranges)):
+ diff_k = diff_by_value_bins[k]
+ mae_k = (np.sum(np.abs(diff_k)) / diff_k.size)
+ values.append(int(mae_k/bin_ranges[k][1] * 100.0))
+
+ print('MAE: ', diff_k.size, bin_ranges[k], mae_k)
+
+ return np.array(values), bin_edges
+
+
+if __name__ == "__main__":
+ nn = SRCNN()
+ nn.run('matchup_filename')