Package 'sigminer.prediction'

Construct A Batch of Keras Models

Description

Construct A Batch of Keras Models

Usage

batch_modeling_and_fitting(data_list, param_combination, ...)

Arguments

data_list	A list containing predictor and label matrix of training data and test data. Please use prepare_data to generate this.
param_combination	A parameter matrix/data.frame with each row representing the parameters for run Keras model once. Column names should indicate parameter names and should be same as in modeling function. base::expand.grid() may be very useful to generate it.
...	Other arguments passing to modeling_and_fitting.

Value

a tibble.

Examples

load(system.file("extdata", "wang2020-input.RData",
  package = "sigminer.prediction", mustWork = TRUE
))
dat_list <- prepare_data(expo_all,
  col_to_vars = c(paste0("Sig", 1:5), paste0("AbsSig", 1:5)),
  col_to_label = "enrich_sig",
  label_names = paste0("Sig", 1:5)
)
pc <- expand.grid(
  c(10, 20, 50, 100),
  c(0, 0.1, 0.2, 0.3, 0.4, 0.5),
  c(10, 20, 50, 100),
  c(0, 0.1, 0.2, 0.3, 0.4, 0.5)
)
colnames(pc) <- c(
  "first_layer_unit", "second_layer_drop_rate",
  "third_layer_unit", "fourth_layer_drop_rate"
)

# Just use 2 rows for illustration
batch_res <- batch_modeling_and_fitting(dat_list, param_combination = pc %>% head(2))
batch_res

tidy(batch_res)

---

Copy Model File

Description

It is usefully when your result model file is stored in temp directory and you want to keep it.

Usage

copy_model(model_file, dest)

Arguments

model_file	A file path to the model file.
dest	The destination file path.

Value

Nothing

---

List Current Available Trained Keras Models

Description

List Current Available Trained Keras Models

Usage

list_trained_models()

Value

A tibble containing summary models.

Examples

list_trained_models()

---

Load Trained Models

Description

Load Trained Models

Usage

load_trained_model(x)

Arguments

x	A subset from list_trained_models.

Value

A (list of) Keras model.

Examples

z <- list_trained_models() %>%
  head(1) %>%
  load_trained_model()
z

---

Create 5-layer Keras Model and Fitting Datasets

Description

Create 5-layer Keras Model and Fitting Datasets

Usage

modeling_and_fitting(
  data_list,
  first_layer_unit,
  second_layer_drop_rate,
  third_layer_unit,
  fourth_layer_drop_rate,
  epochs = 30,
  batch_size = 16,
  validation_split = 0.2,
  validation_data = NULL,
  test_split = NULL,
  first_layer_activation = "relu",
  third_layer_activation = "relu",
  fifth_layer_activation = "softmax",
  loss = "categorical_crossentropy",
  optimizer = optimizer_rmsprop(),
  metrics = c("accuracy"),
  model_file = tempfile(pattern = "keras_model", tmpdir = file.path(tempdir(),
    "sigminer.pred"), fileext = ".h5"),
  test_mode = FALSE
)

Arguments

data_list	A list containing predictor and label matrix of training data and test data. Please use prepare_data to generate this.
first_layer_unit	Positive integer, dimensionality of the output space for the first layer.
second_layer_drop_rate	Float between 0 and 1. Fraction of the input units to drop for the second layer.
third_layer_unit	Positive integer, dimensionality of the output space for the third layer.
fourth_layer_drop_rate	Float between 0 and 1. Fraction of the input units to drop for the fourth layer.
epochs	Number of epochs to train the model, default is 30.
batch_size	Integer or NULL. Number of samples per gradient update. If unspecified, batch_size will default to 16.
validation_split	Float between 0 and 1. Fraction of the training data to be used as validation data. The model will set apart this fraction of the training data, will not train on it, and will evaluate the loss and any model metrics on this data at the end of each epoch. The validation data is selected from the last samples in the x and y data provided, before shuffling.
validation_data	Data on which to evaluate the loss and any model metrics at the end of each epoch. The model will not be trained on this data. This could be a list (x_val, y_val) or a list (x_val, y_val, val_sample_weights). validation_data will override validation_split.
test_split	Float between 0 and 1. Fraction of the all data to be used as test data. If not set, it will be auto-calculated from input data. This value is used for calculating total accuracy.
first_layer_activation	activation function for the first layer, default is "relu".
third_layer_activation	activation function for the third layer, default is "relu".
fifth_layer_activation	activation function for the fifth layer, default is "softmax".
loss	String (name of objective function), objective function or a keras$losses$Loss subclass instance. An objective function is any callable with the signature loss = fn(y_true, y_pred), where y_true = ground truth values with shape = ⁠[batch_size, d0, .. dN]⁠, except sparse loss functions such as sparse categorical crossentropy where shape = ⁠[batch_size, d0, .. dN-1]⁠. y_pred = predicted values with shape = ⁠[batch_size, d0, .. dN]⁠. It returns a weighted loss float tensor. If a custom Loss instance is used and reduction is set to NULL, return value has the shape ⁠[batch_size, d0, .. dN-1]⁠ i.e. per-sample or per-timestep loss values; otherwise, it is a scalar. If the model has multiple outputs, you can use a different loss on each output by passing a dictionary or a list of losses. The loss value that will be minimized by the model will then be the sum of all individual losses, unless loss_weights is specified.
optimizer	String (name of optimizer) or optimizer instance. For most models, this defaults to "rmsprop"
metrics	List of metrics to be evaluated by the model during training and testing. Each of this can be a string (name of a built-in function), function or a keras$metrics$Metric class instance. See ?tf$keras$metrics. Typically you will use metrics=list('accuracy'). A function is any callable with the signature result = fn(y_true, y_pred). To specify different metrics for different outputs of a multi-output model, you could also pass a dictionary, such as metrics=list(output_a = 'accuracy', output_b = c('accuracy', 'mse')). You can also pass a list to specify a metric or a list of metrics for each output, such as metrics=list(list('accuracy'), list('accuracy', 'mse')) or metrics=list('accuracy', c('accuracy', 'mse')). When you pass the strings 'accuracy' or 'acc', this is converted to one of tf.keras.metrics.BinaryAccuracy, tf.keras.metrics.CategoricalAccuracy, tf.keras.metrics.SparseCategoricalAccuracy based on the loss function used and the model output shape. A similar conversion is done for the strings 'crossentropy' and 'ce'.
model_file	file path to save the model file in hdf5 format. Default use a temp file path, the path will be stored in returned data. You can load the model with keras::load_model_hdf5().
test_mode	Default is FALSE, if TRUE, print the input parameters from the user and exit.

Value

a tibble.

Examples

load(system.file("extdata", "wang2020-input.RData",
  package = "sigminer.prediction", mustWork = TRUE
))
dat_list <- prepare_data(expo_all,
  col_to_vars = c(paste0("Sig", 1:5), paste0("AbsSig", 1:5)),
  col_to_label = "enrich_sig",
  label_names = paste0("Sig", 1:5)
)
res <- modeling_and_fitting(dat_list, 20, 0, 20, 0.1)
res$history[[1]] %>% plot()

## Load model and predict
model <- load_model_hdf5(res$model_file)

model %>% predict_classes(dat_list$x_train[1, , drop = FALSE])
model %>% predict_proba(dat_list$x_train[1, , drop = FALSE])

---

Prepare Training and Test Dataset

Description

Prepare Training and Test Dataset

Usage

prepare_data(
  data,
  col_to_vars,
  col_to_label,
  label_names,
  seed = 1234,
  test_split = 0.2
)

Arguments

data	A data.frame.
col_to_vars	A character vector specifying the predictive columns.
col_to_label	A column indicating the labels/classes.
label_names	Label/class names. The order is important. For example, "a", "b", "c" will be transformed to 0, 1, 2.
seed	Random seed, default is 1234.
test_split	A fraction of samples to treated as test dataset, default is 0.2.

Value

a list containing x_train, y_train, x_test, y_test datasets.

Examples

load(system.file("extdata", "wang2020-input.RData",
  package = "sigminer.prediction", mustWork = TRUE
))
dat_list <- prepare_data(expo_all,
  col_to_vars = c(paste0("Sig", 1:5), paste0("AbsSig", 1:5)),
  col_to_label = "enrich_sig",
  label_names = paste0("Sig", 1:5)
)
str(dat_list)

---

Tidy Modeling Result

Description

Tidy Modeling Result

Usage

tidy(x)

Arguments

x	A result tibble from either modeling_and_fitting or batch_modeling_and_fitting.

Value

a tibble

---