Introducing Keras in R
Dr Juan H Klopper
library(keras)
library(readr)
Introduction
Keras is a high-level deep learning network application programming interface. It uses Google’s TensorFlow architecture as back-end. It has become a very popular interface. So much so that version 1.9 of TensorFlow now includes Keras.
For the installation of Keras, follow the instruction at https://keras.rstudio.com/
This post introduces some of the basics of Keras. The dataset is from the UCI machine learning repository and is available at https://archive.ics.uci.edu/ml/datasets/cardiotocography#
Importing the data
The dataset on the UC Irvine servers is in .xls format and contains three sheets. For the purposes of this post, the data has been selected to include \(21\) feature variables and is available as a .csv file at https://github.com/juanklopper/MachineLearningDataSets . The feature variables are described on the UC Irvine website. The observations comprise \(2,126\) fetal cardiotocograms (CTGs) automatically processed and the respective diagnostic features measured. The feature variable is a fetal state coded as 1 = normal, 2 = suspect, and 3 = pathologic, making this a multi-class classification problem.
# Importing the data file as a tibble
df <- read_csv("CTG.csv")## Parsed with column specification:
## cols(
## .default = col_integer(),
## MSTV = col_double(),
## MLTV = col_double()
## )## See spec(...) for full column specifications.df## # A tibble: 2,126 x 22
## LB AC FM UC ASTV MSTV ALTV MLTV DL DS DP Width
## <int> <int> <int> <int> <int> <dbl> <int> <dbl> <int> <int> <int> <int>
## 1 120 0 0 0 73 0.5 43 2.4 0 0 0 64
## 2 132 4 0 4 17 2.1 0 10.4 2 0 0 130
## 3 133 2 0 5 16 2.1 0 13.4 2 0 0 130
## 4 134 2 0 6 16 2.4 0 23 2 0 0 117
## 5 132 4 0 5 16 2.4 0 19.9 0 0 0 117
## 6 134 1 0 10 26 5.9 0 0 9 0 2 150
## 7 134 1 0 9 29 6.3 0 0 6 0 2 150
## 8 122 0 0 0 83 0.5 6 15.6 0 0 0 68
## 9 122 0 0 1 84 0.5 5 13.6 0 0 0 68
## 10 122 0 0 3 86 0.3 6 10.6 0 0 0 68
## # ... with 2,116 more rows, and 10 more variables: Min <int>, Max <int>,
## # Nmax <int>, Nzeros <int>, Mode <int>, Mean <int>, Median <int>,
## # Variance <int>, Tendency <int>, NSP <int>Transforming the data
Keras can manage data in a variety of forms. This post transforms the data into a matrix (without column headers). The result is stored in the computer variable data.
# Convert the data.frame to a matrix
data <- as.matrix(df)
# Remove variable names
dimnames(data) <- NULLIn order to create a model for this multi-class classification problem, the target variable sample space must be converted form 1,2,3 to 0,1,2. This is done through simple broadcasting, subtracting \(1\) from each value. Note that there are \(22\) column in total and that the target variable is in column \(22\).
# Convert target variable values from 1, 2, 3 to 0, 1, 2
data[, 22] <- as.numeric(data[, 22]) - 1 # BroadcastingThe data has to be split into a training and a test set. This can be done through indexing, creating two indices, 1 and 2. The test data is set to comprise \(30\)% of the \(2,126\) samples.
# Split for train and test data
set.seed(123)
indx <- sample(2,
nrow(data),
replace = TRUE,
prob = c(0.7, 0.3)) # Makes index with values 1 and 2The first \(21\) columns containing the feature variables are separated into two computer variables (according to the random index created above).
x_train <- data[indx == 1, 1:21] # Take rows with index = 1
x_test <- data[indx == 2, 1:21]A separate computer variable is created to hold the ground-truth (actual) feature values of the test set for later use.
y_test_actual <- data[indx == 2, 22]The feature variables of the training and test sets can be one-hot-encoded using the Keras function to_categorical().
# Using similar indices to correspond to the training and test set
y_train <-to_categorical(data[indx == 1, 22])
y_test <- to_categorical(data[indx == 2, 22])The sample spaces of the feature variables are not scaled equally. The code chunk below calculates the mean and the standard deviation of the training set and then normalizes both the training and test sets using the mean and standard deviation of the training set (it is important to use the statistics from the training set only!).
mean_train <- apply(x_train,
2,
mean)
std_train <- apply(x_train,
2,
sd)
x_train <- scale(x_train,
center = mean_train,
scale = std_train)
x_test <- scale(x_test,
center = mean_train,
scale = std_train)Creating a model
The model is created in the code chunk below. It comprises a simple sequential layout. Note the use of a pipeline to add new layers. The first is a densely connected layer with 22 nodes. It uses L2 regularization to combat overfitting. The rectified linear unit, relu, activation function is used for this layer. Finally the input dimension is set to the number of feature variables in the data. Deeper layers infer the dimensions from prior layers and are not specified.
The next layer is a dropout layer also employed to combat overfitting and drops an 0.2 fraction of weights.
The next layer is a densely connected layer with 12 nodes and also uses L2 regularization.
The final layer comprises 3 nodes to equal the number of target classes and uses softmax activation so as to provide three probability outputs (summing to \(1\)).
The summary() function summarizes the model and shows a total of 799 trainable parameters.
# Creating the model
model <- keras_model_sequential()
model %>%
layer_dense(units = 22,
kernel_regularizer = regularizer_l2(0.001),
activation = "relu",
input_shape = c(21)) %>%
layer_dropout(rate = 0.2) %>%
layer_dense(units = 12,
kernel_regularizer = regularizer_l2(0.001),
activation = "relu") %>%
layer_dense(units = 3,
activation = "softmax")
summary(model)## ___________________________________________________________________________
## Layer (type) Output Shape Param #
## ===========================================================================
## dense_1 (Dense) (None, 22) 484
## ___________________________________________________________________________
## dropout_1 (Dropout) (None, 22) 0
## ___________________________________________________________________________
## dense_2 (Dense) (None, 12) 276
## ___________________________________________________________________________
## dense_3 (Dense) (None, 3) 39
## ===========================================================================
## Total params: 799
## Trainable params: 799
## Non-trainable params: 0
## ___________________________________________________________________________Compiling the model
Before fitting, the model requires compilation. The loss function, optimizer, and metrics are specified during this step. In this example categorical cross-entropy is used as the loss function (since this is a multi-class classification problem). A standard ADAM optimizer is used and accuracy is used as the metric.
# Compiling the model
model %>% compile(loss = "categorical_crossentropy",
optimizer = "adam",
metrics = c("accuracy"))Fitting the data
The training set is used to fit the compiled model. In addition a validation set is created during the training and is set to comprise a fraction of 0.2 of the training data.
The fitted model is saved in a computer variable named history. Twenty epochs are run, with a mini-batch size of 64.
When using Keras in RStudio, two live plots are created in the Viewer tab. The top shows the loss values for the training and validation sets. The bottom plot shows the accuracy of the two sets.
history <- model %>%
fit(x_train,
y_train,
epoch = 20,
batch_size = 64,
validation_split = 0.2)A simple plot can be created to show the loss and the accuracy over the epochs.
plot(history)
Model evaluation
The test feature and target sets can be used to evaluate the model. The results show the overall loss and accuracy.
model %>%
evaluate(x_test,
y_test)## $loss
## [1] 0.3452709
##
## $acc
## [1] 0.8858954A confusion matrix can be created. A computer variable is created to store the predicted classes. This is used with the initially saved ground-truth values, y_test_actual.
pred <- model %>%
predict_classes(x_test)
table(Predicted = pred,
Actual = y_test_actual)## Actual
## Predicted 0 1 2
## 0 466 22 7
## 1 29 59 13
## 2 0 1 34The predict_proba() function creates probabilities for each of the three classes in each of the test cases. The case with the highest probability is chosen as the predicted target class.
The probabilities, the predicted values, and the ground-truth values can be printed.
prob <- model %>%
predict_proba(x_test)
cbind(prob,
pred,
y_test_actual)## pred y_test_actual
## [1,] 0.9946132302 4.244180e-03 1.142515e-03 0 0
## [2,] 0.9957030416 3.870453e-03 4.265072e-04 0 0
## [3,] 0.9989113808 9.737037e-04 1.148821e-04 0 0
## [4,] 0.6105054617 2.957885e-01 9.370606e-02 0 2
## [5,] 0.5076344013 4.565351e-01 3.583047e-02 0 1
## [6,] 0.9068153501 3.616634e-02 5.701829e-02 0 0
## [7,] 0.9385101199 5.736294e-02 4.127037e-03 0 0
## [8,] 0.0035599803 5.001696e-01 4.962704e-01 1 2
## [9,] 0.2990410626 4.007707e-01 3.001882e-01 1 2
## [10,] 0.0093117626 3.446683e-01 6.460199e-01 2 2
## [11,] 0.9600453973 2.434310e-02 1.561147e-02 0 0
## [12,] 0.9944580793 1.808195e-03 3.733691e-03 0 0
## [13,] 0.9996280670 2.767871e-04 9.512861e-05 0 0
## [14,] 0.9986776710 6.067843e-04 7.154683e-04 0 0
## [15,] 0.9862101674 1.373777e-02 5.209009e-05 0 0
## [16,] 0.8967483044 4.100140e-02 6.225029e-02 0 1
## [17,] 0.5924333334 3.868577e-01 2.070885e-02 0 0
## [18,] 0.6327170134 3.461554e-01 2.112767e-02 0 0
## [19,] 0.9817512631 1.734049e-02 9.082819e-04 0 0
## [20,] 0.8535372019 1.428750e-01 3.587928e-03 0 0
## [21,] 0.9555196166 4.249474e-02 1.985634e-03 0 0
## [22,] 0.9555196166 4.249474e-02 1.985634e-03 0 0
## [23,] 0.7822526097 2.108396e-01 6.907783e-03 0 0
## [24,] 0.9751938581 2.285185e-02 1.954290e-03 0 0
## [25,] 0.9966660142 3.240776e-03 9.312194e-05 0 0
## [26,] 0.9793024659 2.009699e-02 6.005044e-04 0 0
## [27,] 0.9029899836 7.446974e-02 2.254033e-02 0 0
## [28,] 0.3233386278 6.038444e-01 7.281701e-02 1 0
## [29,] 0.1983830780 6.699204e-01 1.316965e-01 1 1
## [30,] 0.9953442216 3.189051e-03 1.466726e-03 0 0
## [31,] 0.9923777580 3.191918e-03 4.430381e-03 0 0
## [32,] 0.9918956161 2.182844e-03 5.921557e-03 0 0
## [33,] 0.9945689440 2.902349e-03 2.528640e-03 0 0
## [34,] 0.9587801099 3.182514e-02 9.394782e-03 0 0
## [35,] 0.9990273714 5.686298e-04 4.039680e-04 0 0
## [36,] 0.9905387163 5.312071e-03 4.149152e-03 0 0
## [37,] 0.4318903983 5.376837e-01 3.042598e-02 1 1
## [38,] 0.3729335368 5.981412e-01 2.892526e-02 1 0
## [39,] 0.6069157720 3.830176e-01 1.006665e-02 0 0
## [40,] 0.6802836061 3.130305e-01 6.685804e-03 0 0
## [41,] 0.3168718517 6.470736e-01 3.605450e-02 1 1
## [42,] 0.5749207735 4.119066e-01 1.317258e-02 0 1
## [43,] 0.9557338357 4.308387e-02 1.182283e-03 0 0
## [44,] 0.6088119149 3.646314e-01 2.655671e-02 0 0
## [45,] 0.3358251750 5.875630e-01 7.661179e-02 1 0
## [46,] 0.9585456848 3.759498e-02 3.859360e-03 0 0
## [47,] 0.9925540090 3.869544e-03 3.576440e-03 0 0
## [48,] 0.9932615161 3.643183e-03 3.095299e-03 0 0
## [49,] 0.9976849556 1.575907e-03 7.391734e-04 0 0
## [50,] 0.9969246984 2.302149e-03 7.731682e-04 0 0
## [51,] 0.9699724913 1.537431e-02 1.465322e-02 0 0
## [52,] 0.7059506774 2.851403e-01 8.909035e-03 0 1
## [53,] 0.9523783922 4.560680e-02 2.014744e-03 0 0
## [54,] 0.6304410100 3.574274e-01 1.213160e-02 0 1
## [55,] 0.4862834215 4.790252e-01 3.469137e-02 0 1
## [56,] 0.8737565279 7.136793e-02 5.487554e-02 0 0
## [57,] 0.9063270688 5.278996e-02 4.088295e-02 0 0
## [58,] 0.8995088339 7.450714e-02 2.598403e-02 0 0
## [59,] 0.9289557338 5.943097e-02 1.161330e-02 0 0
## [60,] 0.9428604841 4.261258e-02 1.452699e-02 0 0
## [61,] 0.8660569787 1.045610e-01 2.938210e-02 0 0
## [62,] 0.9462656379 4.217701e-02 1.155726e-02 0 0
## [63,] 0.9728584290 2.116586e-02 5.975688e-03 0 0
## [64,] 0.9541261196 2.727651e-02 1.859744e-02 0 0
## [65,] 0.0594416223 4.077340e-01 5.328244e-01 2 1
## [66,] 0.9936516285 5.493534e-03 8.548059e-04 0 0
## [67,] 0.6352056265 3.416610e-01 2.313334e-02 0 0
## [68,] 0.9437072277 5.454602e-02 1.746782e-03 0 0
## [69,] 0.1758356988 7.591144e-01 6.504980e-02 1 1
## [70,] 0.9990903139 5.759816e-04 3.337461e-04 0 0
## [71,] 0.9992102385 5.639229e-04 2.258055e-04 0 0
## [72,] 0.9839651585 1.187759e-02 4.157321e-03 0 0
## [73,] 0.9845935702 1.209834e-02 3.308191e-03 0 0
## [74,] 0.9813636541 1.503401e-02 3.602366e-03 0 0
## [75,] 0.9989180565 7.019015e-04 3.800825e-04 0 0
## [76,] 0.6321661472 1.926125e-01 1.752214e-01 0 0
## [77,] 0.9898495674 7.640533e-03 2.509793e-03 0 0
## [78,] 0.9117094874 5.761445e-02 3.067602e-02 0 0
## [79,] 0.9492081404 3.558499e-02 1.520678e-02 0 0
## [80,] 0.2686593533 6.290653e-01 1.022754e-01 1 0
## [81,] 0.1257064492 7.928936e-01 8.139995e-02 1 1
## [82,] 0.0716950968 8.201494e-01 1.081556e-01 1 1
## [83,] 0.0333527811 7.006555e-01 2.659917e-01 1 2
## [84,] 0.0461961403 7.044098e-01 2.493940e-01 1 1
## [85,] 0.0603893623 6.993693e-01 2.402415e-01 1 1
## [86,] 0.1407079697 6.390989e-01 2.201931e-01 1 1
## [87,] 0.0727012530 8.134191e-01 1.138797e-01 1 1
## [88,] 0.1702809334 7.561743e-01 7.354470e-02 1 1
## [89,] 0.1507759839 6.649730e-01 1.842510e-01 1 1
## [90,] 0.0533846281 7.862008e-01 1.604145e-01 1 1
## [91,] 0.0254378710 7.280799e-01 2.464822e-01 1 2
## [92,] 0.1521775275 6.722727e-01 1.755498e-01 1 2
## [93,] 0.0583377741 5.846394e-01 3.570228e-01 1 2
## [94,] 0.0556026436 7.077438e-01 2.366536e-01 1 1
## [95,] 0.0598907322 6.497920e-01 2.903173e-01 1 1
## [96,] 0.1296324879 5.851674e-01 2.852001e-01 1 1
## [97,] 0.0556026436 7.077438e-01 2.366536e-01 1 1
## [98,] 0.0128272679 4.085220e-01 5.786507e-01 2 2
## [99,] 0.0375081673 5.544078e-01 4.080840e-01 1 1
## [100,] 0.0170907490 5.153626e-01 4.675466e-01 1 2
## [101,] 0.3156418502 5.976015e-01 8.675661e-02 1 1
## [102,] 0.8648421168 1.044750e-01 3.068291e-02 0 0
## [103,] 0.0335743316 4.919313e-01 4.744944e-01 1 1
## [104,] 0.2393676937 5.033392e-01 2.572931e-01 1 1
## [105,] 0.2143775374 5.857572e-01 1.998653e-01 1 1
## [106,] 0.2028132081 6.566969e-01 1.404900e-01 1 1
## [107,] 0.9129726291 8.542473e-02 1.602590e-03 0 1
## [108,] 0.3765766919 5.622922e-01 6.113122e-02 1 1
## [109,] 0.9980504513 1.528312e-03 4.212672e-04 0 0
## [110,] 0.9328349829 3.846312e-02 2.870191e-02 0 0
## [111,] 0.9930530190 6.789001e-03 1.579631e-04 0 0
## [112,] 0.9950557947 4.278098e-03 6.660487e-04 0 0
## [113,] 0.0865156576 6.782562e-01 2.352281e-01 1 1
## [114,] 0.9116876125 8.429428e-02 4.018206e-03 0 0
## [115,] 0.8997327685 9.895904e-02 1.308259e-03 0 1
## [116,] 0.1847029775 7.677968e-01 4.750022e-02 1 1
## [117,] 0.1823750138 7.634642e-01 5.416092e-02 1 1
## [118,] 0.0608429164 7.475461e-01 1.916110e-01 1 1
## [119,] 0.2057192028 5.909380e-01 2.033428e-01 1 1
## [120,] 0.5184186697 4.036295e-01 7.795181e-02 0 1
## [121,] 0.3345478177 6.132525e-01 5.219968e-02 1 1
## [122,] 0.9975070357 1.985936e-03 5.070229e-04 0 0
## [123,] 0.9977810979 1.698236e-03 5.205946e-04 0 0
## [124,] 0.9791579843 1.584144e-02 5.000591e-03 0 0
## [125,] 0.6997651458 2.673199e-01 3.291504e-02 0 1
## [126,] 0.4437370896 5.130575e-01 4.320537e-02 1 1
## [127,] 0.1063375324 6.584853e-01 2.351772e-01 1 1
## [128,] 0.1013321653 6.594985e-01 2.391694e-01 1 1
## [129,] 0.7416284084 2.090933e-01 4.927829e-02 0 0
## [130,] 0.9952998161 3.820368e-03 8.797766e-04 0 0
## [131,] 0.9904044867 9.421412e-03 1.741607e-04 0 0
## [132,] 0.8379716873 9.171889e-02 7.030935e-02 0 0
## [133,] 0.0875819921 7.696933e-01 1.427248e-01 1 2
## [134,] 0.0768293887 7.516568e-01 1.715139e-01 1 2
## [135,] 0.0960843861 7.856069e-01 1.183087e-01 1 2
## [136,] 0.5324246287 3.997658e-01 6.780953e-02 0 0
## [137,] 0.1150670573 6.307629e-01 2.541700e-01 1 1
## [138,] 0.3738407195 5.559827e-01 7.017656e-02 1 1
## [139,] 0.9976691604 1.612609e-03 7.182509e-04 0 0
## [140,] 0.9998708963 1.270653e-04 1.988592e-06 0 0
## [141,] 0.9989489913 1.026048e-03 2.497618e-05 0 0
## [142,] 0.9967353940 2.669113e-03 5.955341e-04 0 0
## [143,] 0.9998045564 1.926111e-04 2.841539e-06 0 0
## [144,] 0.9244148135 7.394641e-02 1.638722e-03 0 0
## [145,] 0.5680181980 4.135713e-01 1.841054e-02 0 0
## [146,] 0.9948362112 4.371514e-03 7.923628e-04 0 0
## [147,] 0.9939331412 5.544113e-03 5.227861e-04 0 0
## [148,] 0.9951636791 3.974712e-03 8.616654e-04 0 0
## [149,] 0.9999305010 6.938024e-05 9.089291e-08 0 0
## [150,] 0.9896849394 1.027333e-02 4.181439e-05 0 0
## [151,] 0.9996556044 3.417405e-04 2.653682e-06 0 0
## [152,] 0.3401744068 5.281306e-01 1.316949e-01 1 1
## [153,] 0.6450244784 3.164808e-01 3.849463e-02 0 0
## [154,] 0.2133942991 6.502897e-01 1.363160e-01 1 1
## [155,] 0.9730226398 2.626808e-02 7.093193e-04 0 0
## [156,] 0.8763501644 1.163653e-01 7.284492e-03 0 0
## [157,] 0.1402041018 6.535615e-01 2.062344e-01 1 1
## [158,] 0.1214878187 6.786115e-01 1.999008e-01 1 1
## [159,] 0.3201095462 5.111445e-01 1.687460e-01 1 1
## [160,] 0.5936678648 3.474869e-01 5.884514e-02 0 0
## [161,] 0.9938488603 5.812957e-03 3.382635e-04 0 0
## [162,] 0.9790705442 1.488439e-02 6.045063e-03 0 0
## [163,] 0.9617049694 2.779923e-02 1.049584e-02 0 0
## [164,] 0.9869900346 8.764497e-03 4.245424e-03 0 0
## [165,] 0.9938619137 3.437175e-03 2.700910e-03 0 0
## [166,] 0.9963746667 2.488534e-03 1.136859e-03 0 0
## [167,] 0.9998644590 1.225050e-04 1.304795e-05 0 0
## [168,] 0.9517211318 2.244540e-02 2.583347e-02 0 0
## [169,] 0.9966758490 2.429656e-03 8.945564e-04 0 0
## [170,] 0.9916114211 4.252193e-03 4.136372e-03 0 0
## [171,] 0.9843852520 6.487157e-03 9.127617e-03 0 0
## [172,] 0.9724671841 1.560307e-02 1.192974e-02 0 0
## [173,] 0.9938613176 3.401181e-03 2.737522e-03 0 0
## [174,] 0.9950028062 4.273271e-03 7.238546e-04 0 0
## [175,] 0.2403135598 6.602942e-01 9.939216e-02 1 1
## [176,] 0.2713558376 6.618637e-01 6.678046e-02 1 1
## [177,] 0.8707512617 1.256901e-01 3.558635e-03 0 0
## [178,] 0.7704301476 2.225057e-01 7.064186e-03 0 0
## [179,] 0.9944291115 5.505589e-03 6.528121e-05 0 0
## [180,] 0.9868996143 1.295667e-02 1.437411e-04 0 0
## [181,] 0.9983608127 1.617665e-03 2.155549e-05 0 0
## [182,] 0.9758110046 2.152236e-02 2.666678e-03 0 0
## [183,] 0.8837068081 9.885719e-02 1.743603e-02 0 0
## [184,] 0.9976710677 2.287273e-03 4.157765e-05 0 0
## [185,] 0.8128573895 1.770328e-01 1.010977e-02 0 0
## [186,] 0.9984518290 1.320236e-03 2.279540e-04 0 0
## [187,] 0.9499034882 2.690217e-02 2.319443e-02 0 0
## [188,] 0.9997014403 2.869736e-04 1.154546e-05 0 0
## [189,] 0.7293695807 2.616891e-01 8.941315e-03 0 1
## [190,] 0.4382502437 5.018061e-01 5.994360e-02 1 1
## [191,] 0.5988578796 3.615295e-01 3.961264e-02 0 0
## [192,] 0.9027338028 9.171381e-02 5.552403e-03 0 0
## [193,] 0.9561963081 4.096594e-02 2.837714e-03 0 0
## [194,] 0.9181483388 7.246776e-02 9.383909e-03 0 0
## [195,] 0.9886816740 1.109521e-02 2.231199e-04 0 0
## [196,] 0.8409093618 1.232113e-01 3.587936e-02 0 1
## [197,] 0.0127941594 3.484311e-01 6.387747e-01 2 2
## [198,] 0.0121275224 3.467333e-01 6.411392e-01 2 2
## [199,] 0.0149604548 3.281161e-01 6.569235e-01 2 2
## [200,] 0.9958360195 3.591972e-03 5.720004e-04 0 0
## [201,] 0.9935392737 5.074734e-03 1.385997e-03 0 0
## [202,] 0.9922786951 5.488335e-03 2.232977e-03 0 0
## [203,] 0.5115620494 4.712490e-03 4.837254e-01 0 1
## [204,] 0.1007133648 8.066176e-01 9.266910e-02 1 1
## [205,] 0.3024724126 4.765079e-01 2.210197e-01 1 1
## [206,] 0.9659783244 3.361333e-02 4.083739e-04 0 0
## [207,] 0.5111232400 4.317421e-01 5.713465e-02 0 0
## [208,] 0.9694795609 2.955890e-02 9.616034e-04 0 0
## [209,] 0.0034954683 1.234489e-02 9.841596e-01 2 2
## [210,] 0.0029173852 1.758885e-02 9.794938e-01 2 2
## [211,] 0.9861667752 1.002725e-02 3.805879e-03 0 0
## [212,] 0.0032920893 5.800344e-01 4.166735e-01 1 2
## [213,] 0.0045720544 4.458303e-01 5.495976e-01 2 2
## [214,] 0.9880685210 9.535951e-03 2.395565e-03 0 0
## [215,] 0.9014328718 8.163715e-02 1.692995e-02 0 0
## [216,] 0.1153108925 7.110460e-01 1.736431e-01 1 1
## [217,] 0.0860366896 7.164306e-01 1.975327e-01 1 1
## [218,] 0.3826472163 5.367161e-01 8.063657e-02 1 0
## [219,] 0.9170072675 6.435521e-02 1.863754e-02 0 0
## [220,] 0.7604343891 1.986604e-01 4.090516e-02 0 1
## [221,] 0.9743638039 1.769380e-02 7.942312e-03 0 0
## [222,] 0.9748268127 2.248135e-02 2.691848e-03 0 0
## [223,] 0.9916445613 8.147191e-03 2.082654e-04 0 0
## [224,] 0.9848967791 1.472854e-02 3.746508e-04 0 0
## [225,] 0.9984390140 1.295698e-03 2.653611e-04 0 0
## [226,] 0.9950646758 2.970001e-03 1.965262e-03 0 0
## [227,] 0.9799695611 1.426084e-02 5.769598e-03 0 0
## [228,] 0.0246965177 4.746855e-01 5.006180e-01 2 2
## [229,] 0.8853633404 8.477002e-02 2.986660e-02 0 0
## [230,] 0.0063231937 3.862889e-01 6.073880e-01 2 2
## [231,] 0.0101700826 4.117244e-01 5.781055e-01 2 2
## [232,] 0.0057351887 3.560935e-01 6.381713e-01 2 2
## [233,] 0.1078022197 5.040556e-01 3.881421e-01 1 1
## [234,] 0.0901106298 5.656865e-01 3.442028e-01 1 1
## [235,] 0.1410257518 5.795802e-01 2.793940e-01 1 1
## [236,] 0.9084775448 6.274326e-02 2.877925e-02 0 0
## [237,] 0.9667897820 2.606218e-02 7.148076e-03 0 0
## [238,] 0.6622475386 2.542230e-01 8.352948e-02 0 0
## [239,] 0.3468110859 5.886673e-01 6.452170e-02 1 0
## [240,] 0.6838400960 2.516633e-01 6.449666e-02 0 0
## [241,] 0.8708392978 1.083979e-01 2.076272e-02 0 0
## [242,] 0.9742374420 2.375664e-02 2.005994e-03 0 0
## [243,] 0.1222690940 7.875699e-01 9.016102e-02 1 2
## [244,] 0.0988205075 8.172884e-01 8.389106e-02 1 2
## [245,] 0.2605476081 6.442823e-01 9.517001e-02 1 0
## [246,] 0.2389867604 6.679773e-01 9.303587e-02 1 1
## [247,] 0.2302153260 7.049338e-01 6.485089e-02 1 0
## [248,] 0.9927971959 5.808152e-03 1.394696e-03 0 0
## [249,] 0.1733602434 4.888800e-01 3.377599e-01 1 0
## [250,] 0.3215554655 4.180627e-01 2.603818e-01 1 0
## [251,] 0.9177879691 7.028638e-02 1.192556e-02 0 0
## [252,] 0.9240206480 6.500679e-02 1.097255e-02 0 0
## [253,] 0.8954051733 9.026837e-02 1.432636e-02 0 0
## [254,] 0.5410959125 3.699223e-01 8.898176e-02 0 0
## [255,] 0.8421931863 1.409894e-01 1.681735e-02 0 0
## [256,] 0.8501386642 1.355689e-01 1.429248e-02 0 0
## [257,] 0.9006125331 7.722507e-02 2.216242e-02 0 0
## [258,] 0.8968515992 8.520940e-02 1.793901e-02 0 0
## [259,] 0.9470688105 4.028854e-02 1.264265e-02 0 0
## [260,] 0.9700415730 2.456225e-02 5.396224e-03 0 0
## [261,] 0.8176038861 1.495853e-01 3.281073e-02 0 0
## [262,] 0.5445602536 3.392498e-01 1.161899e-01 0 0
## [263,] 0.8182618022 1.539263e-01 2.781190e-02 0 0
## [264,] 0.9538686275 4.163557e-02 4.495744e-03 0 0
## [265,] 0.9431629181 5.177576e-02 5.061335e-03 0 0
## [266,] 0.8909900188 1.035360e-01 5.473981e-03 0 0
## [267,] 0.8966137767 9.856080e-02 4.825467e-03 0 0
## [268,] 0.9606555700 3.501626e-02 4.328181e-03 0 0
## [269,] 0.9307549000 6.066662e-02 8.578409e-03 0 0
## [270,] 0.9558337331 3.955253e-02 4.613707e-03 0 0
## [271,] 0.9879252911 1.176189e-02 3.128471e-04 0 0
## [272,] 0.9426952004 5.547627e-02 1.828525e-03 0 0
## [273,] 0.9962698221 2.296463e-03 1.433771e-03 0 0
## [274,] 0.9947899580 4.026917e-03 1.183100e-03 0 0
## [275,] 0.9971267581 1.627217e-03 1.246124e-03 0 0
## [276,] 0.9941396713 4.417037e-03 1.443383e-03 0 0
## [277,] 0.9939395189 4.884763e-03 1.175744e-03 0 0
## [278,] 0.9975212216 2.021617e-03 4.571405e-04 0 0
## [279,] 0.9955321550 3.721032e-03 7.468192e-04 0 0
## [280,] 0.9940278530 4.872230e-03 1.099925e-03 0 0
## [281,] 0.9940914512 4.829365e-03 1.079257e-03 0 0
## [282,] 0.0883715749 5.952474e-01 3.163810e-01 1 1
## [283,] 0.9934925437 5.412867e-03 1.094584e-03 0 0
## [284,] 0.9468175173 4.245045e-02 1.073207e-02 0 0
## [285,] 0.9884329438 9.226723e-03 2.340293e-03 0 0
## [286,] 0.9169901013 7.400070e-02 9.009217e-03 0 0
## [287,] 0.8559524417 1.075426e-01 3.650497e-02 0 0
## [288,] 0.9801515937 1.286636e-02 6.981934e-03 0 0
## [289,] 0.9420903325 4.518501e-02 1.272460e-02 0 0
## [290,] 0.9726402760 2.155261e-02 5.807147e-03 0 0
## [291,] 0.9920042157 6.047540e-03 1.948265e-03 0 0
## [292,] 0.9903056622 7.465779e-03 2.228552e-03 0 0
## [293,] 0.9973721504 1.970913e-03 6.569216e-04 0 0
## [294,] 0.9922372699 5.233911e-03 2.528766e-03 0 0
## [295,] 0.9973473549 1.882555e-03 7.700191e-04 0 0
## [296,] 0.9995361567 3.971287e-04 6.674812e-05 0 0
## [297,] 0.9993046522 5.838773e-04 1.114675e-04 0 0
## [298,] 0.9974426031 1.171327e-03 1.386074e-03 0 0
## [299,] 0.9818115830 1.584918e-02 2.339165e-03 0 0
## [300,] 0.9949808717 4.342409e-03 6.766678e-04 0 0
## [301,] 0.9922015667 6.084617e-03 1.713786e-03 0 0
## [302,] 0.9634786844 2.747167e-02 9.049711e-03 0 0
## [303,] 0.9709773064 2.202905e-02 6.993627e-03 0 0
## [304,] 0.9944139719 2.834156e-03 2.751850e-03 0 0
## [305,] 0.9963256717 2.423275e-03 1.251055e-03 0 0
## [306,] 0.9971542358 1.897343e-03 9.484756e-04 0 0
## [307,] 0.9949910045 3.158845e-03 1.850131e-03 0 0
## [308,] 0.9942477942 3.392602e-03 2.359605e-03 0 0
## [309,] 0.9951707721 1.702423e-03 3.126878e-03 0 0
## [310,] 0.9951972365 1.763575e-03 3.039205e-03 0 0
## [311,] 0.9960755706 3.714927e-03 2.094903e-04 0 0
## [312,] 0.9945736527 5.346143e-03 8.021080e-05 0 0
## [313,] 0.9981878400 1.093286e-03 7.188182e-04 0 0
## [314,] 0.9131416082 6.232435e-02 2.453408e-02 0 0
## [315,] 0.9132091999 6.926809e-02 1.752266e-02 0 0
## [316,] 0.9567824006 3.505373e-02 8.163829e-03 0 0
## [317,] 0.9986844659 8.839444e-04 4.315180e-04 0 0
## [318,] 0.9992396832 5.076256e-04 2.526470e-04 0 0
## [319,] 0.9978837371 1.346048e-03 7.702903e-04 0 0
## [320,] 0.9998471737 9.984910e-05 5.292664e-05 0 0
## [321,] 0.9998421669 1.000862e-04 5.770605e-05 0 0
## [322,] 0.9997139573 1.862102e-04 9.981610e-05 0 0
## [323,] 0.9987842441 8.145191e-04 4.012064e-04 0 0
## [324,] 0.9984186888 8.886481e-04 6.927048e-04 0 0
## [325,] 0.9994885921 2.773369e-04 2.339573e-04 0 0
## [326,] 0.9984301925 9.945779e-04 5.752685e-04 0 0
## [327,] 0.9988184571 7.246886e-04 4.568403e-04 0 0
## [328,] 0.9988921285 7.316397e-04 3.762841e-04 0 0
## [329,] 0.9986640215 8.248054e-04 5.112003e-04 0 0
## [330,] 0.9977934361 1.543350e-03 6.632537e-04 0 0
## [331,] 0.9987903237 8.596150e-04 3.500826e-04 0 0
## [332,] 0.9976655245 1.637514e-03 6.969621e-04 0 0
## [333,] 0.9971547127 1.804883e-03 1.040407e-03 0 0
## [334,] 0.9988658428 6.485285e-04 4.856842e-04 0 0
## [335,] 0.9985694885 8.837685e-04 5.467418e-04 0 0
## [336,] 0.9987881780 8.590689e-04 3.528414e-04 0 0
## [337,] 0.9941903353 4.131001e-03 1.678728e-03 0 0
## [338,] 0.9965316653 2.588031e-03 8.802885e-04 0 0
## [339,] 0.9983534813 1.203686e-03 4.428776e-04 0 0
## [340,] 0.9896509647 5.054859e-03 5.294103e-03 0 0
## [341,] 0.9966276288 1.131964e-03 2.240371e-03 0 0
## [342,] 0.9959751964 2.727675e-03 1.297129e-03 0 0
## [343,] 0.9982536435 1.203754e-03 5.426673e-04 0 0
## [344,] 0.9974406958 1.901018e-03 6.582637e-04 0 0
## [345,] 0.9965672493 2.502887e-03 9.298897e-04 0 0
## [346,] 0.9918244481 6.985650e-03 1.189897e-03 0 0
## [347,] 0.9904862046 7.991496e-03 1.522317e-03 0 0
## [348,] 0.9929732084 5.624833e-03 1.402028e-03 0 0
## [349,] 0.9663905501 2.867925e-02 4.930213e-03 0 0
## [350,] 0.9768800735 2.043954e-02 2.680374e-03 0 0
## [351,] 0.8293679357 1.353954e-01 3.523668e-02 0 0
## [352,] 0.8905023932 1.031373e-01 6.360317e-03 0 0
## [353,] 0.8308190703 1.592882e-01 9.892728e-03 0 0
## [354,] 0.8709914088 1.207822e-01 8.226413e-03 0 0
## [355,] 0.2085499465 6.745694e-01 1.168806e-01 1 1
## [356,] 0.9243522882 6.386212e-02 1.178566e-02 0 0
## [357,] 0.4081416726 4.803571e-01 1.115012e-01 1 1
## [358,] 0.5539919138 3.691403e-01 7.686780e-02 0 1
## [359,] 0.8377544880 1.347175e-01 2.752806e-02 0 0
## [360,] 0.8072927594 1.377794e-01 5.492787e-02 0 0
## [361,] 0.9468358755 4.161515e-02 1.154903e-02 0 0
## [362,] 0.9397232533 4.826556e-02 1.201123e-02 0 0
## [363,] 0.9954784513 3.304878e-03 1.216699e-03 0 0
## [364,] 0.9829289913 1.044920e-02 6.621774e-03 0 0
## [365,] 0.9992871881 4.055091e-04 3.073574e-04 0 0
## [366,] 0.9997081161 1.557005e-04 1.362560e-04 0 0
## [367,] 0.9998615980 7.836044e-05 6.010814e-05 0 0
## [368,] 0.9998457432 8.754891e-05 6.658693e-05 0 0
## [369,] 0.9979084730 9.041824e-04 1.187365e-03 0 0
## [370,] 0.9950365424 3.106738e-03 1.856745e-03 0 0
## [371,] 0.9991294742 6.282711e-04 2.422224e-04 0 0
## [372,] 0.9983635545 1.084827e-03 5.516761e-04 0 0
## [373,] 0.9982483387 1.273683e-03 4.780115e-04 0 0
## [374,] 0.9995106459 3.528408e-04 1.365330e-04 0 0
## [375,] 0.9996126294 2.450232e-04 1.423536e-04 0 0
## [376,] 0.9998617172 7.117748e-05 6.706911e-05 0 0
## [377,] 0.9995818734 2.247294e-04 1.934108e-04 0 0
## [378,] 0.9995954633 2.136018e-04 1.909163e-04 0 0
## [379,] 0.9989503026 6.714240e-04 3.783823e-04 0 0
## [380,] 0.9993601441 5.185326e-04 1.212791e-04 0 0
## [381,] 0.9995880723 2.624315e-04 1.495734e-04 0 0
## [382,] 0.9930981994 4.139750e-03 2.762027e-03 0 0
## [383,] 0.9921436310 3.056774e-03 4.799638e-03 0 0
## [384,] 0.9973369241 1.720666e-03 9.423254e-04 0 0
## [385,] 0.9435349703 5.097874e-02 5.486344e-03 0 0
## [386,] 0.9455053806 4.656192e-02 7.932645e-03 0 0
## [387,] 0.9994375110 3.580417e-04 2.045015e-04 0 0
## [388,] 0.9985103011 1.226445e-03 2.631996e-04 0 0
## [389,] 0.9906214476 5.863685e-03 3.514841e-03 0 0
## [390,] 0.9768383503 9.845918e-03 1.331571e-02 0 0
## [391,] 0.9401748776 2.536296e-02 3.446213e-02 0 0
## [392,] 0.9469978213 2.478054e-02 2.822161e-02 0 0
## [393,] 0.9460963607 2.619843e-02 2.770513e-02 0 0
## [394,] 0.9479073882 3.501092e-02 1.708166e-02 0 0
## [395,] 0.9517282248 1.644185e-02 3.182990e-02 0 0
## [396,] 0.5670106411 2.101122e-01 2.228772e-01 0 1
## [397,] 0.9969239831 2.496356e-03 5.796361e-04 0 0
## [398,] 0.9930035472 5.815507e-03 1.180905e-03 0 0
## [399,] 0.9994142056 3.808026e-04 2.050322e-04 0 0
## [400,] 0.9962106943 2.350953e-03 1.438355e-03 0 0
## [401,] 0.9836909771 8.403317e-03 7.905716e-03 0 0
## [402,] 0.9906043410 7.186947e-03 2.208767e-03 0 0
## [403,] 0.9841690063 1.245114e-02 3.379869e-03 0 0
## [404,] 0.7931916118 1.833867e-01 2.342180e-02 0 1
## [405,] 0.4252495766 5.234703e-01 5.128011e-02 1 1
## [406,] 0.5326750875 4.289375e-01 3.838745e-02 0 1
## [407,] 0.8169877529 1.708617e-01 1.215048e-02 0 0
## [408,] 0.6949657798 2.737139e-01 3.132031e-02 0 0
## [409,] 0.8244033456 1.637945e-01 1.180222e-02 0 0
## [410,] 0.5798700452 3.668619e-01 5.326807e-02 0 0
## [411,] 0.9738771915 2.435599e-02 1.766835e-03 0 0
## [412,] 0.9801436067 1.847816e-02 1.378278e-03 0 0
## [413,] 0.9841684699 1.491614e-02 9.155048e-04 0 0
## [414,] 0.9668808579 3.035178e-02 2.767457e-03 0 0
## [415,] 0.8989087939 9.510463e-02 5.986555e-03 0 0
## [416,] 0.8296200633 1.597077e-01 1.067225e-02 0 0
## [417,] 0.8683230281 1.243923e-01 7.284658e-03 0 0
## [418,] 0.9837585092 1.534411e-02 8.973566e-04 0 0
## [419,] 0.9774871469 2.118851e-02 1.324337e-03 0 0
## [420,] 0.9465352297 5.136812e-02 2.096585e-03 0 0
## [421,] 0.6242234707 3.550915e-01 2.068510e-02 0 0
## [422,] 0.3941429257 5.737400e-01 3.211706e-02 1 1
## [423,] 0.6726093292 3.074330e-01 1.995770e-02 0 0
## [424,] 0.8055576086 1.862352e-01 8.207058e-03 0 0
## [425,] 0.5198024511 4.652123e-01 1.498528e-02 0 1
## [426,] 0.8137383461 1.777438e-01 8.517849e-03 0 0
## [427,] 0.8137383461 1.777438e-01 8.517849e-03 0 0
## [428,] 0.7821717858 2.088832e-01 8.945055e-03 0 0
## [429,] 0.3134058416 6.194718e-01 6.712227e-02 1 0
## [430,] 0.2698401809 6.630375e-01 6.712240e-02 1 1
## [431,] 0.3113425076 6.540752e-01 3.458233e-02 1 0
## [432,] 0.3377572596 6.260560e-01 3.618676e-02 1 0
## [433,] 0.5212445259 4.568331e-01 2.192230e-02 0 1
## [434,] 0.8630036116 1.314294e-01 5.567057e-03 0 0
## [435,] 0.9542632103 4.458925e-02 1.147601e-03 0 1
## [436,] 0.9923806787 7.526127e-03 9.316811e-05 0 0
## [437,] 0.9974663258 2.505783e-03 2.798743e-05 0 0
## [438,] 0.9890088439 1.083251e-02 1.586227e-04 0 0
## [439,] 0.9742648005 2.552119e-02 2.140524e-04 0 0
## [440,] 0.9569677711 3.212009e-02 1.091217e-02 0 0
## [441,] 0.9256697297 5.602974e-02 1.830053e-02 0 0
## [442,] 0.9957367182 3.488957e-03 7.743287e-04 0 0
## [443,] 0.9833526611 1.356609e-02 3.081217e-03 0 0
## [444,] 0.9602867365 1.366874e-02 2.604451e-02 0 2
## [445,] 0.9936066866 4.953608e-03 1.439750e-03 0 2
## [446,] 0.9854024649 1.068963e-02 3.907931e-03 0 0
## [447,] 0.9842029810 1.180351e-02 3.993531e-03 0 0
## [448,] 0.9913937449 7.827726e-03 7.785216e-04 0 0
## [449,] 0.9984440207 1.482569e-03 7.347670e-05 0 0
## [450,] 0.9931138158 6.478969e-03 4.072647e-04 0 0
## [451,] 0.9860532880 1.037173e-02 3.574979e-03 0 0
## [452,] 0.9858953357 1.019071e-02 3.913905e-03 0 0
## [453,] 0.9806112051 1.424599e-02 5.142773e-03 0 0
## [454,] 0.9719894528 2.136554e-02 6.645095e-03 0 0
## [455,] 0.9922958612 6.115066e-03 1.589139e-03 0 0
## [456,] 0.9709166288 2.501692e-02 4.066526e-03 0 0
## [457,] 0.9585212469 3.178084e-02 9.697857e-03 0 0
## [458,] 0.6982750297 2.570725e-01 4.465242e-02 0 0
## [459,] 0.9754847288 1.986419e-02 4.651032e-03 0 0
## [460,] 0.3337758780 6.058441e-01 6.037999e-02 1 1
## [461,] 0.5668935776 4.026136e-01 3.049276e-02 0 0
## [462,] 0.8488253355 1.435755e-01 7.599175e-03 0 0
## [463,] 0.9071385860 8.545934e-02 7.402169e-03 0 0
## [464,] 0.8859189749 1.092348e-01 4.846218e-03 0 0
## [465,] 0.9982703924 1.544046e-03 1.854939e-04 0 0
## [466,] 0.9765439034 1.004208e-02 1.341406e-02 0 0
## [467,] 0.8547118306 4.035308e-03 1.412528e-01 0 0
## [468,] 0.9056912065 6.313759e-02 3.117123e-02 0 0
## [469,] 0.9696570635 2.593948e-02 4.403509e-03 0 0
## [470,] 0.8565143347 1.004476e-01 4.303803e-02 0 0
## [471,] 0.9702833295 2.021386e-02 9.502866e-03 0 0
## [472,] 0.9899114966 9.458638e-03 6.299083e-04 0 0
## [473,] 0.9967072606 3.146677e-03 1.461264e-04 0 0
## [474,] 0.8031152487 1.792622e-01 1.762253e-02 0 1
## [475,] 0.8395195603 1.173584e-01 4.312201e-02 0 0
## [476,] 0.9280153513 3.237688e-02 3.960784e-02 0 0
## [477,] 0.5714246035 3.200283e-01 1.085471e-01 0 0
## [478,] 0.9459975958 2.769990e-02 2.630258e-02 0 0
## [479,] 0.9996560812 2.559263e-04 8.799271e-05 0 0
## [480,] 0.9975389242 1.792189e-03 6.688950e-04 0 0
## [481,] 0.9993144274 5.616259e-04 1.239697e-04 0 0
## [482,] 0.9991574287 6.798696e-04 1.626849e-04 0 0
## [483,] 0.9808343053 1.427176e-02 4.893897e-03 0 0
## [484,] 0.9799730182 1.323370e-02 6.793336e-03 0 0
## [485,] 0.9815077186 1.393271e-02 4.559588e-03 0 0
## [486,] 0.9888012409 6.285841e-03 4.912877e-03 0 0
## [487,] 0.9805933237 1.059327e-02 8.813425e-03 0 0
## [488,] 0.9939076304 4.023108e-03 2.069318e-03 0 0
## [489,] 0.9751130342 1.444761e-02 1.043935e-02 0 0
## [490,] 0.9785631299 1.076453e-02 1.067235e-02 0 0
## [491,] 0.9962279797 2.584764e-03 1.187238e-03 0 0
## [492,] 0.9367485642 6.331694e-03 5.691977e-02 0 0
## [493,] 0.9948587418 4.504364e-03 6.368738e-04 0 0
## [494,] 0.9984897375 9.807374e-04 5.295168e-04 0 0
## [495,] 0.9977319241 1.971450e-03 2.966464e-04 0 0
## [496,] 0.9974363446 2.178066e-03 3.856071e-04 0 0
## [497,] 0.9963592887 2.843907e-03 7.967221e-04 0 0
## [498,] 0.9828010798 4.993487e-03 1.220545e-02 0 0
## [499,] 0.0312971510 4.763508e-02 9.210678e-01 2 2
## [500,] 0.9983112812 1.661367e-03 2.740046e-05 0 0
## [501,] 0.9993748069 5.933380e-04 3.190096e-05 0 0
## [502,] 0.9993116856 6.429988e-04 4.530193e-05 0 0
## [503,] 0.9979087114 1.989551e-03 1.016622e-04 0 0
## [504,] 0.9952594638 4.443581e-03 2.969451e-04 0 0
## [505,] 0.9968085885 2.872583e-03 3.188741e-04 0 0
## [506,] 0.9854907990 1.200875e-02 2.500479e-03 0 0
## [507,] 0.8174410462 1.470970e-01 3.546202e-02 0 0
## [508,] 0.9368434548 5.047225e-02 1.268422e-02 0 0
## [509,] 0.9627342820 2.974058e-02 7.525118e-03 0 0
## [510,] 0.0634697750 3.378932e-01 5.986370e-01 2 2
## [511,] 0.9854635596 7.979914e-03 6.556528e-03 0 0
## [512,] 0.9941835999 5.108268e-03 7.081328e-04 0 0
## [513,] 0.9892235994 9.895463e-03 8.810423e-04 0 0
## [514,] 0.9786096811 1.446211e-02 6.928130e-03 0 0
## [515,] 0.8510530591 1.068563e-01 4.209073e-02 0 0
## [516,] 0.9867485762 1.218105e-02 1.070474e-03 0 0
## [517,] 0.9216077924 6.905039e-02 9.341723e-03 0 0
## [518,] 0.9537630081 3.983181e-02 6.405124e-03 0 0
## [519,] 0.9130146503 6.826049e-02 1.872475e-02 0 0
## [520,] 0.9316355586 5.549116e-02 1.287328e-02 0 0
## [521,] 0.8756276369 9.479479e-02 2.957762e-02 0 0
## [522,] 0.9342575669 5.678232e-02 8.960113e-03 0 0
## [523,] 0.9817813635 1.597041e-02 2.248136e-03 0 0
## [524,] 0.9910059571 8.512247e-03 4.817910e-04 0 0
## [525,] 0.9916961789 7.659047e-03 6.447458e-04 0 0
## [526,] 0.0042280434 2.826668e-02 9.675052e-01 2 2
## [527,] 0.0093870275 2.632994e-02 9.642830e-01 2 2
## [528,] 0.0014953598 2.952580e-02 9.689789e-01 2 2
## [529,] 0.0006672847 6.557018e-02 9.337626e-01 2 2
## [530,] 0.0007263742 7.795458e-02 9.213191e-01 2 2
## [531,] 0.8595961332 4.677278e-02 9.363104e-02 0 0
## [532,] 0.5290926695 5.617261e-02 4.147348e-01 0 2
## [533,] 0.3789827824 6.474225e-02 5.562750e-01 2 2
## [534,] 0.0282995068 7.022043e-02 9.014800e-01 2 2
## [535,] 0.0373731293 1.389386e-01 8.236883e-01 2 2
## [536,] 0.9761699438 1.299941e-02 1.083061e-02 0 0
## [537,] 0.9385123253 3.050936e-02 3.097834e-02 0 0
## [538,] 0.9057211280 2.560685e-02 6.867212e-02 0 0
## [539,] 0.9708442688 4.865281e-03 2.429039e-02 0 2
## [540,] 0.9429650903 5.477452e-02 2.260316e-03 0 0
## [541,] 0.5030334592 4.235660e-01 7.340050e-02 0 0
## [542,] 0.2542088926 6.107273e-01 1.350639e-01 1 0
## [543,] 0.2450504154 6.149352e-01 1.400144e-01 1 0
## [544,] 0.3168603182 5.826703e-01 1.004694e-01 1 0
## [545,] 0.8005273938 1.839530e-01 1.551961e-02 0 0
## [546,] 0.9484370351 3.411219e-02 1.745069e-02 0 0
## [547,] 0.7607189417 1.628084e-01 7.647271e-02 0 0
## [548,] 0.7708999515 1.646945e-01 6.440552e-02 0 0
## [549,] 0.7568395734 1.667915e-01 7.636896e-02 0 0
## [550,] 0.7564714551 1.852469e-01 5.828171e-02 0 0
## [551,] 0.7761991620 1.585954e-01 6.520544e-02 0 0
## [552,] 0.7296259999 2.001267e-01 7.024723e-02 0 0
## [553,] 0.6956227422 2.353008e-01 6.907641e-02 0 0
## [554,] 0.8801913261 9.354566e-02 2.626306e-02 0 0
## [555,] 0.8477496505 1.097466e-01 4.250375e-02 0 0
## [556,] 0.8443407416 1.129301e-01 4.272922e-02 0 0
## [557,] 0.9910655618 8.316826e-03 6.175521e-04 0 0
## [558,] 0.9922689795 7.320643e-03 4.104498e-04 0 0
## [559,] 0.9881401062 1.060013e-02 1.259674e-03 0 0
## [560,] 0.9918295741 7.436239e-03 7.341058e-04 0 0
## [561,] 0.9923819900 6.943973e-03 6.740986e-04 0 0
## [562,] 0.9747970104 2.177958e-02 3.423349e-03 0 0
## [563,] 0.9526139498 3.969532e-02 7.690790e-03 0 0
## [564,] 0.9061971307 7.251900e-02 2.128385e-02 0 0
## [565,] 0.1283903122 1.466303e-01 7.249794e-01 2 2
## [566,] 0.8215309978 7.210388e-02 1.063651e-01 0 2
## [567,] 0.8726754785 1.138770e-01 1.344744e-02 0 0
## [568,] 0.8896721005 1.037519e-01 6.576041e-03 0 0
## [569,] 0.9092342854 7.841029e-02 1.235551e-02 0 0
## [570,] 0.9354841709 5.241186e-02 1.210399e-02 0 0
## [571,] 0.9767504334 1.880524e-02 4.444332e-03 0 0
## [572,] 0.8980415463 2.279243e-02 7.916604e-02 0 2
## [573,] 0.0351083614 4.325327e-03 9.605663e-01 2 2
## [574,] 0.6640928984 2.336596e-01 1.022475e-01 0 0
## [575,] 0.8356272578 1.430400e-01 2.133279e-02 0 0
## [576,] 0.7407515049 2.252995e-01 3.394899e-02 0 0
## [577,] 0.2603251338 4.150984e-01 3.245764e-01 1 0
## [578,] 0.9365696311 4.143633e-02 2.199409e-02 0 0
## [579,] 0.9237341285 5.010906e-02 2.615679e-02 0 0
## [580,] 0.0230104998 2.668961e-01 7.100933e-01 2 2
## [581,] 0.0170269217 2.876945e-01 6.952785e-01 2 2
## [582,] 0.0114708515 1.074055e-01 8.811236e-01 2 2
## [583,] 0.0566911362 1.731514e-01 7.701575e-01 2 2
## [584,] 0.0049361465 6.612650e-02 9.289374e-01 2 2
## [585,] 0.9040650725 8.374257e-02 1.219237e-02 0 0
## [586,] 0.8343775272 1.428618e-01 2.276068e-02 0 0
## [587,] 0.9687823653 2.703587e-02 4.181836e-03 0 0
## [588,] 0.9309754968 5.680745e-02 1.221702e-02 0 0
## [589,] 0.9035285711 7.872650e-02 1.774492e-02 0 0
## [590,] 0.9618052840 3.213252e-02 6.062207e-03 0 0
## [591,] 0.8465824127 1.058208e-01 4.759686e-02 0 0
## [592,] 0.8845206499 8.282141e-02 3.265789e-02 0 0
## [593,] 0.5084320903 2.806293e-01 2.109386e-01 0 0
## [594,] 0.4646073878 3.041759e-01 2.312167e-01 0 0
## [595,] 0.6024591923 2.208762e-01 1.766646e-01 0 0
## [596,] 0.7584817410 8.900270e-02 1.525156e-01 0 0
## [597,] 0.8038570285 7.603873e-02 1.201043e-01 0 0
## [598,] 0.0191824622 1.198369e-01 8.609806e-01 2 2
## [599,] 0.0185793731 1.052508e-01 8.761699e-01 2 2
## [600,] 0.0054881182 1.609918e-01 8.335200e-01 2 2
## [601,] 0.0252095815 1.516907e-01 8.230997e-01 2 2
## [602,] 0.0391297676 1.755425e-01 7.853277e-01 2 2
## [603,] 0.8330864310 8.314547e-02 8.376809e-02 0 0
## [604,] 0.8645236492 6.495204e-02 7.052428e-02 0 0
## [605,] 0.8411332965 8.612918e-02 7.273750e-02 0 0
## [606,] 0.8797325492 5.792337e-02 6.234401e-02 0 0
## [607,] 0.8197347522 1.045271e-01 7.573816e-02 0 0
## [608,] 0.3046348691 3.992483e-01 2.961168e-01 1 0
## [609,] 0.6897583008 2.049394e-01 1.053024e-01 0 0
## [610,] 0.7575643063 1.618158e-01 8.061984e-02 0 0
## [611,] 0.2502494454 4.343514e-01 3.153991e-01 1 0
## [612,] 0.2450092733 5.177214e-01 2.372693e-01 1 0
## [613,] 0.9366919994 4.613210e-02 1.717596e-02 0 0
## [614,] 0.8654675484 9.073854e-02 4.379392e-02 0 0
## [615,] 0.8813753724 9.133087e-02 2.729381e-02 0 0
## [616,] 0.6795225739 2.084380e-01 1.120395e-01 0 0
## [617,] 0.4910447598 3.434515e-01 1.655037e-01 0 0
## [618,] 0.9257774949 3.698128e-02 3.724115e-02 0 0
## [619,] 0.0869316757 6.652297e-01 2.478387e-01 1 0
## [620,] 0.3102424145 5.177749e-01 1.719826e-01 1 0
## [621,] 0.3988372087 4.732268e-01 1.279359e-01 1 0
## [622,] 0.6791277528 2.461433e-01 7.472888e-02 0 0
## [623,] 0.3048119545 5.528204e-01 1.423677e-01 1 0
## [624,] 0.8773652911 9.032656e-02 3.230823e-02 0 0
## [625,] 0.8564434052 1.040799e-01 3.947673e-02 0 0
## [626,] 0.3498455584 3.865705e-01 2.635840e-01 1 0
## [627,] 0.0995591432 6.448596e-01 2.555813e-01 1 0
## [628,] 0.2604725063 4.950401e-01 2.444873e-01 1 0
## [629,] 0.1804920584 5.113162e-01 3.081917e-01 1 0
## [630,] 0.2513023913 5.473306e-01 2.013671e-01 1 0
## [631,] 0.3690575957 5.495579e-01 8.138452e-02 1 1Conclusion
The Keras package in R makes for an easy to use, powerful deep learning platform. Give it a try!