Plot Data

Øystein Olav Skaar

2019-02-04

Plot Data

Enjoy this brief demonstration of the plot data module

First we simulate some data and estimate means and standard deviations

# Create normal distributed data with mean = 0 and standard deviation = 1
Sigma <- matrix(0.25,3,3)
diag(Sigma) <- 1
set.seed(100)
data <- MASS::mvrnorm(n=1000,mu=c(0,5,10), Sigma=Sigma, empirical=TRUE)
colnames(data) <- c("Before","During","After")

mcmc <- bfw::bfw(project.data = data,
                 y = "Before,During,After",
                 saved.steps = 50000,
                 jags.model = "mean",
                 job.title = "Stages of Cheese",
                 jags.seed = 100,
                 silent = TRUE)

# Print output
round(mcmc$summary.MCMC,3)
#>                  Mean Median   Mode   ESS  HDIlo  HDIhi    n
#> mu[1]: Before       0      0  0.002 51201 -0.062  0.063 1000
#> mu[2]: During       5      5  5.000 50000  4.939  5.064 1000
#> mu[3]: After       10     10 10.000 50000  9.938 10.062 1000
#> sigma[1]: Before    1      1  1.000 49354  0.958  1.046 1000
#> sigma[2]: During    1      1  1.000 50000  0.957  1.045 1000
#> sigma[3]: After     1      1  0.997 50000  0.957  1.045 1000
Plot <- bfw::PlotMean(mcmc,
                      run.repeated = TRUE)
ParsePlot(Plot)

Plot the data as repeated measures

plot1

plot1

Lets add some noise

set.seed(101)
noise <- apply(data,2, function (x) x + rbinom(length(x),1,0.7))

noise.mcmc <- bfw::bfw(project.data = noise,
                  y = "Before,During,After",
                  saved.steps = 50000,
                  jags.model = "mean",
                  job.title = "Stages of Cheese",
                  jags.seed = 101,
                  silent = TRUE)

# Print output
round(noise.mcmc$summary.MCMC,3)
#>                    Mean Median   Mode   ESS  HDIlo  HDIhi    n
#> mu[1]: Before     0.713  0.713  0.713 50000  0.641  0.781 1000
#> mu[2]: During     5.686  5.686  5.690 48350  5.618  5.756 1000
#> mu[3]: After     10.686 10.686 10.685 50648 10.617 10.753 1000
#> sigma[1]: Before  1.120  1.119  1.116 50000  1.072  1.170 1000
#> sigma[2]: During  1.116  1.116  1.112 50000  1.068  1.166 1000
#> sigma[3]: After   1.101  1.100  1.097 49233  1.054  1.151 1000
Plot <- bfw::PlotMean(noise.mcmc, 
                      run.repeated = TRUE)
ParsePlot(Plot)

Plot the noise as repeated measures

plot2

plot2

Let’s add a group

combined.data <- as.data.frame(rbind(cbind(data,"Y"), cbind(noise,"X") ), stringsAsFactors=FALSE)
combined.data[,1:3] <- lapply(combined.data[,1:3] , as.numeric)
combined.data[,4] <- as.factor(combined.data[,4])
colnames(combined.data) <- c(colnames(data), "Groups")

combined.data <- bfw::bfw(project.data = combined.data,
                     y = "Before,During,After",
                     x = "Groups",
                     job.title = "Stages of Cheese",
                     saved.steps = 50000,
                     jags.model = "mean",
                     jags.seed = 102,
                     silent = TRUE)

# Print output
round(combined.data$summary.MCMC[, 3:7],3)
#>                                   Mode   ESS  HDIlo  HDIhi    n
#> mu[1]: Before                    0.359 50000  0.309  0.407 2000
#> mu[2]: Before vs. Groups @ X     0.713 50000  0.641  0.779 1000
#> mu[3]: Before vs. Groups @ Y    -0.002 50000 -0.063  0.062 1000
#> mu[4]: During                    5.342 50000  5.293  5.391 2000
#> mu[5]: During vs. Groups @ X     5.683 49103  5.616  5.754 1000
#> mu[6]: During vs. Groups @ Y     4.998 50000  4.938  5.062 1000
#> mu[7]: After                    10.344 50000 10.293 10.390 2000
#> mu[8]: After vs. Groups @ X     10.688 49245 10.618 10.754 1000
#> mu[9]: After vs. Groups @ Y      9.999 50000  9.938 10.063 1000
#> sigma[1]: Before                 1.119 50000  1.086  1.155 2000
#> sigma[2]: Before vs. Groups @ X  1.119 48125  1.071  1.169 1000
#> sigma[3]: Before vs. Groups @ Y  0.998 50000  0.958  1.045 1000
#> sigma[4]: During                 1.112 50000  1.079  1.148 2000
#> sigma[5]: During vs. Groups @ X  1.115 49356  1.068  1.165 1000
#> sigma[6]: During vs. Groups @ Y  0.999 50000  0.957  1.045 1000
#> sigma[7]: After                  1.105 50000  1.072  1.140 2000
#> sigma[8]: After vs. Groups @ X   1.098 50000  1.054  1.150 1000
#> sigma[9]: After vs. Groups @ Y   1.000 50000  0.957  1.045 1000

# Let's also add some colors!
Plot <- bfw::PlotMean(combined.data, 
                      run.split = TRUE, 
                      run.repeated = TRUE,  
                      monochrome = FALSE)
ParsePlot(Plot)

Plot the split data

plot3

plot3