In AlfonsoRReyes/RepDataPeerAssessment1: Reproducible Research Assignment 1. Developed as a package

This is the logic. There are 288 5-min intervals the whle day. We could divide the day in 4: from 00:01 to 06:00, from 06:01 to 12:00, from 12:01 to 18:00, and from 18:01 to 24:00. This will help us to define the clusters in the heatmap to draw some distiction.

library(RepDataPeerAssessment1)

data(intervals)
intervals

Remove the NAs.

ok <- complete.cases(intervals)
ok

intervals.na.not <- intervals[ok,]
intervals.na.not

# summing up the rows looking for any observation that is all zeros. The minimum is 41 steps and maximum for the whole day is 21, 194.

rowSums(intervals.na.not[2:289])
min(rowSums(intervals.na.not[2:289]))
max(rowSums(intervals.na.not[2:289]))

intervals.m <- data.matrix(intervals.na.not)
# intervals.m
str(intervals.m)

head(intervals.m[, 1:2])

cols <- seq(2, 289)
heatmap(intervals.m[, cols], Colv = NA, Rowv = NA)

heatmap(intervals.m[, 2:65], Colv = NA, Rowv = NA)
heatmap(intervals.m[, 66:260], Colv = NA, Rowv = NA)
heatmap(intervals.m[, 261:289], Colv = NA, Rowv = NA)

intervals.na.not

cc <- rainbow(74, start = 0, end = .3)

heatmap(intervals.m[, 2:75], Colv = NA, ColSideColors = cc)

cols <- seq(2, 289)
image(intervals.m[, cols], 
      col = grey(seq(0, 1, length = 256)),
      # col = heat.colors(12)
      )

matplot(y = actbycol, type = 'l', lty = 1)

form <- as.formula(paste(paste(names(actbycol)[- 280],  
                               collapse = ' + '),  
                         'date',  sep = '~'))

xyplot(form,  data = actbycol, outer = TRUE)

library(lattice)

df <- data.frame(time = 1:10,
                 a = cumsum(rnorm(10)),
                 b = cumsum(rnorm(10)),
                 c = cumsum(rnorm(10)))

form <- as.formula(paste(paste(names(df)[- 1],  
                               collapse = ' + '),  
                         'time',  sep = '~'))

xyplot(form,  data = df,  type = 'b',  outer = TRUE)

plot.ts(actbycol)

plot(intervals$date, intervals$`1500`, col = "blue")
points(intervals$date, intervals$`1510`, col = "green")

require(graphics); require(grDevices)
x  <- as.matrix(mtcars)
rc <- rainbow(nrow(x), start = 0, end = .3)
cc <- rainbow(ncol(x), start = 0, end = .3)
hv <- heatmap(x, col = cm.colors(256), scale = "column",
              RowSideColors = rc, ColSideColors = cc, margins = c(5,10),
              xlab = "specification variables", ylab =  "Car Models",
              main = "heatmap(<Mtcars data>, ..., scale = \"column\")")
utils::str(hv) # the two re-ordering index vectors

## no column dendrogram (nor reordering) at all:
heatmap(x, Colv = NA, col = cm.colors(256), scale = "column",
        RowSideColors = rc, margins = c(5,10),
        xlab = "specification variables", ylab =  "Car Models",
        main = "heatmap(<Mtcars data>, ..., scale = \"column\")")

## "no nothing"
heatmap(x, Rowv = NA, Colv = NA, scale = "column",
        main = "heatmap(*, NA, NA) ~= image(t(x))")


round(Ca <- cor(attitude), 2)
symnum(Ca) # simple graphic
heatmap(Ca,               symm = TRUE, margins = c(6,6)) # with reorder()
heatmap(Ca, Rowv = FALSE, symm = TRUE, margins = c(6,6)) # _NO_ reorder()
## slightly artificial with color bar, without and with ordering:
cc <- rainbow(nrow(Ca))
heatmap(Ca, Rowv = FALSE, symm = TRUE, RowSideColors = cc, ColSideColors = cc,
    margins = c(6,6))
heatmap(Ca,     symm = TRUE, RowSideColors = cc, ColSideColors = cc,
    margins = c(6,6))

## For variable clustering, rather use distance based on cor():
symnum( cU <- cor(USJudgeRatings) )

hU <- heatmap(cU, Rowv = FALSE, symm = TRUE, col = topo.colors(16),
             distfun = function(c) as.dist(1 - c), keep.dendro = TRUE)
## The Correlation matrix with same reordering:
round(100 * cU[hU[[1]], hU[[2]]])
## The column dendrogram:
utils::str(hU$Colv)

AlfonsoRReyes/RepDataPeerAssessment1 documentation built on May 5, 2019, 4:53 a.m.