Course8_Project
Rui La
11/15/2016
Prediction Assignment Writeupless
Background
Using devices such as Jawbone Up, Nike FuelBand, and Fitbit it is now possible to collect a large amount of data about personal activity relatively inexpensively. These type of devices are part of the quantified self movement – a group of enthusiasts who take measurements about themselves regularly to improve their health, to find patterns in their behavior, or because they are tech geeks. One thing that people regularly do is quantify how much of a particular activity they do, but they rarely quantify how well they do it. In this project, your goal will be to use data from accelerometers on the belt, forearm, arm, and dumbell of 6 participants. They were asked to perform barbell lifts correctly and incorrectly in 5 different ways. More information is available from the website here: http://groupware.les.inf.puc-rio.br/har (see the section on the Weight Lifting Exercise Dataset).
Load data first
getwd()## [1] "/Users/rula/Larry/Coursera/ML"pml_train = read.csv("./pml-training.csv")
pml_test = read.csv("./pml-testing.csv")
library(caret)check dimension and names of variables
names(pml_train)## [1] "X" "user_name"
## [3] "raw_timestamp_part_1" "raw_timestamp_part_2"
## [5] "cvtd_timestamp" "new_window"
## [7] "num_window" "roll_belt"
## [9] "pitch_belt" "yaw_belt"
## [11] "total_accel_belt" "kurtosis_roll_belt"
## [13] "kurtosis_picth_belt" "kurtosis_yaw_belt"
## [15] "skewness_roll_belt" "skewness_roll_belt.1"
## [17] "skewness_yaw_belt" "max_roll_belt"
## [19] "max_picth_belt" "max_yaw_belt"
## [21] "min_roll_belt" "min_pitch_belt"
## [23] "min_yaw_belt" "amplitude_roll_belt"
## [25] "amplitude_pitch_belt" "amplitude_yaw_belt"
## [27] "var_total_accel_belt" "avg_roll_belt"
## [29] "stddev_roll_belt" "var_roll_belt"
## [31] "avg_pitch_belt" "stddev_pitch_belt"
## [33] "var_pitch_belt" "avg_yaw_belt"
## [35] "stddev_yaw_belt" "var_yaw_belt"
## [37] "gyros_belt_x" "gyros_belt_y"
## [39] "gyros_belt_z" "accel_belt_x"
## [41] "accel_belt_y" "accel_belt_z"
## [43] "magnet_belt_x" "magnet_belt_y"
## [45] "magnet_belt_z" "roll_arm"
## [47] "pitch_arm" "yaw_arm"
## [49] "total_accel_arm" "var_accel_arm"
## [51] "avg_roll_arm" "stddev_roll_arm"
## [53] "var_roll_arm" "avg_pitch_arm"
## [55] "stddev_pitch_arm" "var_pitch_arm"
## [57] "avg_yaw_arm" "stddev_yaw_arm"
## [59] "var_yaw_arm" "gyros_arm_x"
## [61] "gyros_arm_y" "gyros_arm_z"
## [63] "accel_arm_x" "accel_arm_y"
## [65] "accel_arm_z" "magnet_arm_x"
## [67] "magnet_arm_y" "magnet_arm_z"
## [69] "kurtosis_roll_arm" "kurtosis_picth_arm"
## [71] "kurtosis_yaw_arm" "skewness_roll_arm"
## [73] "skewness_pitch_arm" "skewness_yaw_arm"
## [75] "max_roll_arm" "max_picth_arm"
## [77] "max_yaw_arm" "min_roll_arm"
## [79] "min_pitch_arm" "min_yaw_arm"
## [81] "amplitude_roll_arm" "amplitude_pitch_arm"
## [83] "amplitude_yaw_arm" "roll_dumbbell"
## [85] "pitch_dumbbell" "yaw_dumbbell"
## [87] "kurtosis_roll_dumbbell" "kurtosis_picth_dumbbell"
## [89] "kurtosis_yaw_dumbbell" "skewness_roll_dumbbell"
## [91] "skewness_pitch_dumbbell" "skewness_yaw_dumbbell"
## [93] "max_roll_dumbbell" "max_picth_dumbbell"
## [95] "max_yaw_dumbbell" "min_roll_dumbbell"
## [97] "min_pitch_dumbbell" "min_yaw_dumbbell"
## [99] "amplitude_roll_dumbbell" "amplitude_pitch_dumbbell"
## [101] "amplitude_yaw_dumbbell" "total_accel_dumbbell"
## [103] "var_accel_dumbbell" "avg_roll_dumbbell"
## [105] "stddev_roll_dumbbell" "var_roll_dumbbell"
## [107] "avg_pitch_dumbbell" "stddev_pitch_dumbbell"
## [109] "var_pitch_dumbbell" "avg_yaw_dumbbell"
## [111] "stddev_yaw_dumbbell" "var_yaw_dumbbell"
## [113] "gyros_dumbbell_x" "gyros_dumbbell_y"
## [115] "gyros_dumbbell_z" "accel_dumbbell_x"
## [117] "accel_dumbbell_y" "accel_dumbbell_z"
## [119] "magnet_dumbbell_x" "magnet_dumbbell_y"
## [121] "magnet_dumbbell_z" "roll_forearm"
## [123] "pitch_forearm" "yaw_forearm"
## [125] "kurtosis_roll_forearm" "kurtosis_picth_forearm"
## [127] "kurtosis_yaw_forearm" "skewness_roll_forearm"
## [129] "skewness_pitch_forearm" "skewness_yaw_forearm"
## [131] "max_roll_forearm" "max_picth_forearm"
## [133] "max_yaw_forearm" "min_roll_forearm"
## [135] "min_pitch_forearm" "min_yaw_forearm"
## [137] "amplitude_roll_forearm" "amplitude_pitch_forearm"
## [139] "amplitude_yaw_forearm" "total_accel_forearm"
## [141] "var_accel_forearm" "avg_roll_forearm"
## [143] "stddev_roll_forearm" "var_roll_forearm"
## [145] "avg_pitch_forearm" "stddev_pitch_forearm"
## [147] "var_pitch_forearm" "avg_yaw_forearm"
## [149] "stddev_yaw_forearm" "var_yaw_forearm"
## [151] "gyros_forearm_x" "gyros_forearm_y"
## [153] "gyros_forearm_z" "accel_forearm_x"
## [155] "accel_forearm_y" "accel_forearm_z"
## [157] "magnet_forearm_x" "magnet_forearm_y"
## [159] "magnet_forearm_z" "classe"dim(pml_train)## [1] 19622 160table(pml_train$classe)##
## A B C D E
## 5580 3797 3422 3216 3607preprocessing
First, I partitioned the training set
clear up near zero variance, missing values and descriptive fields.
training <- pml_train[, 6:dim(pml_train)[2]]
threshold = dim(training)[1]*0.8
judge90 <- function (x) {
sum(is.na(x))>threshold ||sum(x == "") > threshold
}
goodColumns = !apply(training, 2, judge90)
training = training[,goodColumns]
badCol = nearZeroVar(training, saveMetrics = TRUE)
training = training [, badCol$nzv == FALSE]
training$classe = factor(training$classe)- partitioning the training data set to allow cross-validation
inTrain = createDataPartition(training$classe, p = 0.5)[[1]]
trainset = training[inTrain,]
crossVset = training[-inTrain,]
inTrain = createDataPartition(crossVset$classe, p = 0.7)[[1]]
crossTrain = crossVset [inTrain,]
crossTest = crossVset[-inTrain,]
testing = pml_test[, 6:dim(pml_test)[2]]
testing = testing[, goodColumns]
testing$classe = NA
testing = testing [, badCol$nzv == FALSE]
#mod1 = train(classe~., data = training, method = "rf")
#pred1 = predict(mod1, testing)Now we can train with different model. randomForest() is faster than train(“rf”) Don’t know why. train(method = “rf”) takes forever to run.
library(randomForest)## randomForest 4.6-12## Type rfNews() to see new features/changes/bug fixes.##
## Attaching package: 'randomForest'## The following object is masked from 'package:ggplot2':
##
## marginmod1 =randomForest(classe~., data = training, method = "class")
pre1 = predict (mod1, newdata = crossTrain, type = "class")library(randomForest)## randomForest 4.6-12## Type rfNews() to see new features/changes/bug fixes.pre11 = predict(mod1,crossVset, type = "class" )
acc1 = caret::confusionMatrix(pre11, crossVset$classe)
acc1## $positive
## NULL
##
## $table
## Reference
## Prediction A B C D E
## A 2790 0 0 0 0
## B 0 1898 0 0 0
## C 0 0 1711 0 0
## D 0 0 0 1608 0
## E 0 0 0 0 1803
##
## $overall
## Accuracy Kappa AccuracyLower AccuracyUpper AccuracyNull
## 1.0000000 1.0000000 0.9996240 1.0000000 0.2844037
## AccuracyPValue McnemarPValue
## 0.0000000 NaN
##
## $byClass
## Sensitivity Specificity Pos Pred Value Neg Pred Value Precision
## Class: A 1 1 1 1 1
## Class: B 1 1 1 1 1
## Class: C 1 1 1 1 1
## Class: D 1 1 1 1 1
## Class: E 1 1 1 1 1
## Recall F1 Prevalence Detection Rate Detection Prevalence
## Class: A 1 1 0.2844037 0.2844037 0.2844037
## Class: B 1 1 0.1934760 0.1934760 0.1934760
## Class: C 1 1 0.1744139 0.1744139 0.1744139
## Class: D 1 1 0.1639144 0.1639144 0.1639144
## Class: E 1 1 0.1837920 0.1837920 0.1837920
## Balanced Accuracy
## Class: A 1
## Class: B 1
## Class: C 1
## Class: D 1
## Class: E 1
##
## $mode
## [1] "sens_spec"
##
## $dots
## list()
##
## attr(,"class")
## [1] "confusionMatrix"From the confusionMatrix result we can see the prediction is quite right. Now let’s look at “rpart” method
library(rpart)
library(rpart.plot)## Warning: package 'rpart.plot' was built under R version 3.2.5mod2 = rpart(classe ~., data = training, method = "class")
pre2 = predict (mod2, crossTrain, type = "class")
acc2 = caret:: confusionMatrix(pre2, crossTrain$classe)
rpart.plot(mod2, main = "Classification Tree", extra = 102, under = TRUE, faclen = 0)
acc2## $positive
## NULL
##
## $table
## Reference
## Prediction A B C D E
## A 1733 225 49 71 65
## B 70 821 74 95 143
## C 23 88 965 164 92
## D 101 162 78 741 146
## E 26 33 32 55 817
##
## $overall
## Accuracy Kappa AccuracyLower AccuracyUpper AccuracyNull
## 7.391178e-01 6.691450e-01 7.285605e-01 7.494717e-01 2.843209e-01
## AccuracyPValue McnemarPValue
## 0.000000e+00 9.571560e-59
##
## $byClass
## Sensitivity Specificity Pos Pred Value Neg Pred Value Precision
## Class: A 0.8873528 0.9165989 0.8086794 0.9534490 0.8086794
## Class: B 0.6177577 0.9310469 0.6824605 0.9103424 0.6824605
## Class: C 0.8055092 0.9352848 0.7244745 0.9579195 0.7244745
## Class: D 0.6580817 0.9152011 0.6034202 0.9317497 0.6034202
## Class: E 0.6468725 0.9739565 0.8483904 0.9244836 0.8483904
## Recall F1 Prevalence Detection Rate
## Class: A 0.8873528 0.8461914 0.2843209 0.2522929
## Class: B 0.6177577 0.6484992 0.1934779 0.1195225
## Class: C 0.8055092 0.7628458 0.1744068 0.1404862
## Class: D 0.6580817 0.6295667 0.1639249 0.1078760
## Class: E 0.6468725 0.7340521 0.1838696 0.1189402
## Detection Prevalence Balanced Accuracy
## Class: A 0.3119814 0.9019758
## Class: B 0.1751347 0.7744023
## Class: C 0.1939147 0.8703970
## Class: D 0.1787742 0.7866414
## Class: E 0.1401951 0.8104145
##
## $mode
## [1] "sens_spec"
##
## $dots
## list()
##
## attr(,"class")
## [1] "confusionMatrix"As expected, Random Forest algorithm performed better than Rpart. Accuracy for Random Forest model was 1 compared to 0.74 for rpart model. The random Forest model is choosen. The accuracy of the model is 1. The expected out-of-sample error is estimated at 0. The expected out-of-sample error is calculated as 1 - accuracy for predictions made against the cross-validation set. Our Test data set comprises 20 cases. With an accuracy above 99% on our cross-validation data, we can expect that very few, or none, of the test samples will be missclassified.
Now we cansee the rpart result is not accurate. The accuracy is only 74%. So we will choose “rf” method.
For the final testing data
# predict final outcome levels on the original TESTING data using our mod1 algorithm
finalpre = predict(mod1, testing, type = "class")
finalpre## 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
## B A B A A E D B A A B C B A E E A B B B
## Levels: A B C D E