![mosaic plot mosaic plot](https://haleyjeppson.github.io/ggmosaic/articles/ggmosaic_files/figure-html/variety-1.png)
We've used the Grouping feature to clearly differentiate between each group of servers
![mosaic plot mosaic plot](https://communitybucketdatagraph.s3.us-east-2.amazonaws.com/community/datagraph/2603/m5953ta6dky8f13a0d1ag2moihaoxnp1.png)
The second shows the average memory utilization of every server The first shows the average CPU utilization of every server Describe the nature of the association between two categorical variables. The top left heatmap for example (showing the number of requests per server in us-east-1), contains 2880 different cells/data points, while taking up the space that would normally fit only a sparkline or two, let alone trying to fit a line/area chart in that space Use bar graphs and mosaic plots to compare distributions of categorical data. One that shows the number of requests per server over time (white to blue) and one that shows the average error rate per server over time (green to red)īy using Compact Mode we achieve an incredibly high data resolution. library (ggplot2) example <- read.csv ('EXAMPLE.csv', header TRUE, sep' ') ggplot (data example. Each individual has value 0, 1 or 2 for each position. We have no way to see the behavior of specific serversīy using Mosaic Plot panels, we can construct something like this:įirst, for each region, we've added two heatmaps I want to create a plot with individuals A, B and C in y, position in x (its a position not a value, so I dont want a proportional representation on the graph).
![mosaic plot mosaic plot](https://static.wixstatic.com/media/4c04a2_7ee4b7a40f9a40c5893d9762ed14e6dd~mv2_d_1598_1252_s_2.png)
It is somewhat comparable to the compound bar chart, with the exception that the width of each bar is now determined by the fraction of each horizontal category. We have 50 different servers in 4 different regions, and are interested in seeing for each server, the number of requests, error rate, CPU utilization and memory utilization The mosaic plot can be further subdivided into a specialized variety known as a spine plot, which focuses on just two variables. The mosaic plot, being based on a Chi-square test, does not address significance except by yielding a single, overall, 'omnibus' p-value. Nor does any individual residual value determine significance in this context. Color Black White Red Green Blue Yellow Magenta Cyan Transparency Transparent Semi-Transparent Opaque. The plot does not give us sufficient information to further specify the values of each residual. The following example showcases how a Mosaic Plot can be used to effectively visualize servers stats for a large number of servers There is no mosaic plot procedure available in SPSS, but there is a macro created by Andrew Wheeler, which will generate a mosaic. Color Black White Red Green Blue Yellow Magenta Cyan Transparency Opaque Semi-Transparent Transparent.