I am actually using Tidymodels package on R to study a multi-class classification problem. I have trained several models using Workflow sets, and in my recipe I added a step taken there to replace NA values with a constant. The models that I included in the workflow are:
mlp <-
mlp(hidden_units = tune(), penalty = tune(), epochs = tune()) %>%
set_engine('nnet') %>%
set_mode('classification')
multinom <-
multinom_reg(penalty = tune(), mixture = tune()) %>%
set_engine('glmnet')
rand_forest <-
rand_forest(mtry = tune(), min_n = tune()) %>%
set_engine('ranger') %>%
set_mode('classification')
tabnet <- tabnet(mode="classification", batch_size= 126, virtual_batch_size= 128, epochs= 1,
num_steps = tune(), learn_rate = tune())%>%
set_engine("torch", verbose = TRUE)
For some models I tried a recipe with SMOTE ("themis" package), PCA, and normalisation (all in the same workflow by adding the steps to the original recipe). Training and testing went pretty well, so I tried an ensemble of these models (using the package "stacks"):
tidymodels_prefer()
stack1 <-
stacks() %>%
add_candidates(res_1)
set.seed(2002)
res1_stack <-
stack1 %>%
blend_predictions()
ens <- fit_members(res1_stack)
When I run this last operation (fit_members) I receive this error
Error in UseMethod("filter") :
no applicable method for 'filter' applied to an object of class "NULL"
I figured out, reading this and this on GitHub, that it was because the added step "constantimpute" to the recipe. However, I don't exactly know how can I fix it. Someone can help me?
Thank you very much!!!
Before using the filter function, make sure the table you want to filter is loaded.
Most times we have the the view() function applied and this prevents the table from being loaded into memory for usage.
Related
I have made an object of S4 class SeqExpressionSet with EDASeq which I can then analyse with plotPCA.
However the plotPCA function lacks the ability to fully adjust aesthetics. I was therefore wondering whether it is possible to change the dataset somehow so I can use it with e.g. ggplot2 or a different package that enables more adjustments.
I'm not familiar with EDAseq, but my best guess is that you'd have to do a PCA manually and plot those results. Assuming your object is called my_object and the source code posted here, you can reconstruct the process as follows:
dat <- normCounts(my_object)
dat <- apply(dat, 1, function(y) scale(y, center = TRUE, scale = FALSE))
s <- svd(dat)
df <- data.frame(
PC1 = s$u[, 1], PC2 = s$u[, 2]
)
ggplot(df, aes(PC1, PC2)) +
geom_point()
Note that I haven't tested this code as I don't have example data and was too lazy to install EDAseq.
I have got this dataset. I am trying to do a stacking bar graph with proportions using ggplot for this data:
I am not really sure how to manipulate it into tables first! I know, I just started learning R, two weeks ago and I'm kind of stuck. I made a similar graph before. I attached it here.
I'm not sure if I got your question right, but I'll try to answer it. I see that this is your first question in Stack Overflow, so I'd advise you to post a minimal reproducible example on your next question.
1) "I am not really sure how to manipulate it into tables first!"
Copy the data into an excel file, save it as csv and import into R with base R command.
df <- read.csv('your_data.csv')
2) " do a stacking bar graph with proportions"
Your problem is very similar to the one mentioned in this question. Make sure to check it out, but I've already adapted the code below, see if it works.
library(ggplot2)
library(dplyr)
library(tidyr)
df <- read.csv('your_data.csv')
# Add an id variable for the filled regions and reshape
dfm <- df %>%
mutate(Domain = factor(row_number()) %>%
gather(variable, value, -Domain)
ggplot(dfm, aes(x = variable, y = value, fill = Domain)) +
geom_bar(position = "fill",stat = "identity") +
# or:
# geom_bar(position = position_fill(), stat = "identity"
scale_y_continuous(labels = scales::percent_format())
I am trying to build the BiseNet shown in the figure at "https://github.com/Blaizzy/BiSeNet-Implementation".
When I want to use the GlobalAveragePooling2D() in Keras(tf-backend) to finish the Attention Refined Module in Figure(b), I find the output shape of the GlobalAveragePooling2D() is not suitable for the next convolution.
I checked out many implementation of BiSeNet code in github, however, most of them use AveragePooling2D(size=(1,1)) instead. But AveragePooling2D(size=(1,1)) is completely non-sense.
So I define a lambda layer to do what I want (The selected code is shown as below). The lambda layer works but seems very ugly:
def samesize_globalAveragePooling2D(inputtensor):
# inputtensor shape:(?, 28,28,32)
x = GlobalAveragePooling2D()(inputtensor) # x shape:(?, 32)
divide = tf.divide(inputtensor, inputtensor) # divide shape:(?, 28,28,32)
x2 = x * divide # x2 shape:(?, 28,28,32)
global_pool = Lambda(function=samesize_globalAveragePooling2D)(conv_0)
Hope to get suggestion to make this lambda to be more graceful.
Thanks!
This could be done using a lambda layer on tf.reduce_mean.
tf.keras.layers.Lambda(lambda x: tf.reduce_mean(x, axis=[1, 2], keep_dims=True))
I have the following code to average embeddings for list of item-ids.
(Embedding is trained on review_meta_id_input, and used as look up for pirors_input and for getting average embedding)
review_meta_id_input = tf.keras.layers.Input(shape=(1,), dtype='int32', name='review_meta_id')
priors_input = tf.keras.layers.Input(shape=(None,), dtype='int32', name='priors') # array of ids
item_embedding_layer = tf.keras.layers.Embedding(
input_dim=100, # max number
output_dim=self.item_embedding_size,
name='item')
review_meta_id_embedding = item_embedding_layer(review_meta_id_input)
selected = tf.nn.embedding_lookup(review_meta_id_embedding, priors_input)
non_zero_count = tf.cast(tf.math.count_nonzero(priors_input, axis=1), tf.float32)
embedding_sum = tf.reduce_sum(selected, axis=1)
item_average = tf.math.divide(embedding_sum, non_zero_count)
I also have some feature columns such as..
(I just thought feature_column looked cool, but not many documents to look for..)
kid_youngest_month = feature_column.numeric_column("kid_youngest_month")
kid_age_youngest_buckets = feature_column.bucketized_column(kid_youngest_month, boundaries=[12, 24, 36, 72, 96])
I'd like to define [review_meta_id_iput, priors_input, (tensors from feature_columns)] as an input to keras Model.
something like:
inputs = [review_meta_id_input, priors_input] + feature_layer
model = tf.keras.models.Model(inputs=inputs, outputs=o)
In order to get tensors from feature columns, the closest lead I have now is
fc_to_tensor = {fc: input_layer(features, [fc]) for fc in feature_columns}
from https://github.com/tensorflow/tensorflow/issues/17170
However I'm not sure what the features are in the code.
There's no clear example on https://www.tensorflow.org/api_docs/python/tf/feature_column/input_layer either.
How should I construct the features variable for fc_to_tensor ?
Or is there a way to use keras.layers.Input and feature_column at the same time?
Or is there an alternative than tf.feature_column to do the bucketing as above? then I'll just drop the feature_column for now;
The behavior you desire could be achieved through following steps.
This works in TF 2.0.0-beta1, but may being changed or even simplified in further reseases.
Please check out issue in TensorFlow github repository Unable to use FeatureColumn with Keras Functional API #27416. There you will find the more general example and useful comments about tf.feature_column and Keras Functional API.
Meanwhile, based on the code in your question the input tensor for feature_column could be get like this:
# This you have defined feauture column
kid_youngest_month = feature_column.numeric_column("kid_youngest_month")
kid_age_youngest_buckets = feature_column.bucketized_column(kid_youngest_month, boundaries=[12, 24, 36, 72, 96])
# Then define layer
feature_layer = tf.keras.layers.DenseFeatures(kid_age_youngest_buckets)
# The inputs for DenseFeature layer should be define for each original feature column as dictionary, where
# keys - names of feature columns
# values - tf.keras.Input with shape =(1,), name='name_of_feature_column', dtype - actual type of original column
feature_layer_inputs = {}
feature_layer_inputs['kid_youngest_month'] = tf.keras.Input(shape=(1,), name='kid_youngest_month', dtype=tf.int8)
# Then you can collect inputs of other layers and feature_layer_inputs into one list
inputs=[review_meta_id_input, priors_input, [v for v in feature_layer_inputs.values()]]
# Then define outputs of this DenseFeature layer
feature_layer_outputs = feature_layer(feature_layer_inputs)
# And pass them into other layer like any other
x = tf.keras.layers.Dense(256, activation='relu')(feature_layer_outputs)
# Or maybe concatenate them with outputs from your others layers
combined = tf.keras.layers.concatenate([x, feature_layer_outputs])
#And probably you will finish with last output layer, maybe like this for calssification
o=tf.keras.layers.Dense(classes_number, activation='softmax', name='sequential_output')(combined)
#So you pass to the model:
model_combined = tf.keras.models.Model(inputs=[s_inputs, [v for v in feature_layer_inputs.values()]], outputs=o)
Also note. In model fit() method you should pass info which data sould be used for each input.
One way, if you use tf.data.Dataset, take care that you have used the same names for features in Dataset and for keys in feature_layer_inputs dictionary
Other way use explicite notation like:
model.fit({'review_meta_id_input': review_meta_id_data, 'priors_input': priors_data, 'kid_youngest_month': kid_youngest_month_data},
{'outputs': o},
...
)
I am getting this error message
Error in mx.model.select.layout.train(X, y) : Cannot auto select
array.layout, please specify this parameter
I am trying to create simple model to understand how things work
create data
train.x <- data.matrix(sample(1:100,1000,replace=T) )
colnames(train.x) <- "X"
train.y <- data.matrix(train.x^2)
colnames(train.y) <- "Y"
test.x = data.matrix(sample(1:100,50,replace=T) )
colnames(test.x) <- "X"
test.y <-data.matrix(test.x^2)
colnames(test.y) <- "Y"
test data
mx.set.seed(0)
model <- mx.mlp(train.x, train.y, hidden_node=10, out_node=2,
out_activation="softmax",
num.round=20, array.batch.size=15, learning.rate=0.07, momentum=0.9,
eval.metric=mx.metric.accuracy)
It looks like you're trying to learn an approximation of of the function x --> x ^ 2. Your activation function shouldn't then be "softmax". That would be more appropriate for a classification problem. You could MSE (mean-squared error) or other loss functions more suited to a regression problem.
You might also find this MXNet/R tutorial helpful:
https://mxnet.incubator.apache.org/tutorials/r/fiveMinutesNeuralNetwork.html