機械学習

演習課題

d <- read.csv(file = 'https://stats.dip.jp/01_ds/data/data_cleansing2_dirty_data.csv')

m <- as.matrix(d) 

boxplot(m)

any(m < 0)

## [1] TRUE

any(abs(m) > 3*sd(m))

## [1] TRUE

jj <- 36:46
boxplot(m[, jj])

j <- 37
y <- m[, j]

barplot(height    = y,         
        names.arg = 1:nrow(m))

ii <- 140:231
y <- m[ii, j]

barplot(height    = y,                        
        cex.names = 0.4,            
        names.arg = paste(ii, '\n(', y, ')'))

ii <- 194:206
y <- m[ii, j]

barplot(height    = y,                        
        cex.names = 0.4,                      
        names.arg = paste(ii, '\n(', y, ')'))

m[200,j]

##   V37 
## -0.78

m[m<0]

## [1] -21.00  -0.78  -0.99  -0.68

sigma <- sd(m)

m[abs(m) > 4*sigma]

## [1]  22 -21  11

m.a <- NULL 

for (j in 1:ncol(m))
{
  for (i in 1:nrow(m))
  { 
    if ( m[i, j] < 0 | m[i, j] > 4 * sigma )
    {
      cat('Row:', i, ', Col:', j, ', Value: ', m[i, j], fill = T) 
      m.a <- rbind(m.a, t(c(i, j, m[i, j]))) 
    }
  }
}

## Row: 39 , Col: 1 , Value:  22
## Row: 52 , Col: 1 , Value:  -21
## Row: 91 , Col: 1 , Value:  11
## Row: 200 , Col: 37 , Value:  -0.78
## Row: 521 , Col: 59 , Value:  -0.99
## Row: 242 , Col: 100 , Value:  -0.68

colnames(m.a) <- c('Row','Col','Value')
m.a

##      Row Col  Value
## [1,]  39   1  22.00
## [2,]  52   1 -21.00
## [3,]  91   1  11.00
## [4,] 200  37  -0.78
## [5,] 521  59  -0.99
## [6,] 242 100  -0.68

d <- read.csv(file = 'https://stats.dip.jp/01_ds/data/data_cleansing2_dirty_data_exercise.csv')

library(DT)

datatable(d)

boxplot(d)

any(d < 0)  

## [1] TRUE

jj <- 36:46
boxplot(d[, jj])

j <- 37
y <- d[, j]

barplot(height    = y,         
        names.arg = 1:nrow(d))

ii <- 140:231
y <- d[ii, j]

barplot(height    = y,                        
        cex.names = 0.4,                      
        names.arg = paste(ii, '\n(', y, ')'))

ii <- 194:206
y <- d[ii, j]

barplot(height    = y,                        
        cex.names = 0.4,                     
        names.arg = paste(ii, '\n(', y, ')'))

d[200, j]

## [1] 0.67

d[d < 0]

## [1] -42.00  -0.92  -0.55  -0.03

d.a <- NULL 

for (j in 1:ncol(d))
{
  for (i in 1:nrow(d))
  { 
    if ( d[i, j] < 0 | d[i, j] > 4 * sigma )
    {
      cat('Row:', i, ', Col:', j, ', Value: ', d[i, j], fill = T) 
      d.a <- rbind(d.a, t(c(i, j, d[i, j]))) 
    }
  }
}

## Row: 2 , Col: 1 , Value:  -42
## Row: 42 , Col: 1 , Value:  22
## Row: 99 , Col: 1 , Value:  111
## Row: 531 , Col: 27 , Value:  -0.92
## Row: 675 , Col: 83 , Value:  -0.55
## Row: 124 , Col: 85 , Value:  -0.03