View source: R/Equitable_Transform.R
plotsummary | R Documentation |
Plots of various output from a transform: shows both Equitable and Least squares results Will compare them to the original data and to a signal is it is available Various formats for displaying the dat are used including images, contours and row/column plots with error bars shown (Because many plots are produced you may want to put the plots into a pdf file (foo) using pdf(file=foo)) before and dev.off() after using the function plotsummary )
plotsummary( Td_noise, Td = NULL, Td_old = NULL, row_unit = NULL, col_unit = NULL, z_unit = NULL, yline = 3, yma = 5, fintersect = FALSE, fsquares = FALSE, fpca = TRUE, fave = FALSE, fall = FALSE, inc = NULL, plim = NULL )
Td_noise |
Output from the transform function transformE fro the data to be studied |
Td |
NULL(Default) Output from the transform program for an underlying signal. Allows comparisons with undelrlying signal |
Td_old |
ignore |
row_unit |
name for the row dimension for axis plotting e.g. "Day number" |
col_unit |
name for the row dimension for axis plotting e.g. "Year" |
z_unit |
name for the measured quantity e.g. "Temperature" |
yline |
3 (default) number of lines from image to start ylabel |
yma |
5 (default) distance in from margin to start images |
fintersect |
FALSE (default) TRUE: plots intercept vs 1-slope for different zeroes |
fsquares |
TRUE (default) line p[lots] produced of slope and shift square matrices |
fave |
FALSE (default) TRUE: shows results with errors of performing averaging ion data |
fall |
FALSE (default) TRUE: all "events are used to construct bagplots when finterswect is also TRUE |
inc |
NULL(default) 10 colums plotted : inc when set is the increment in columns that the plots step through |
None
# first create a data set d and create the associated transforms. # In this case d is eg7 with a resolution 3x higher than the lowest #consider putting the graphs into a pdf file by bracketing your #commands beginning with pdf(file="foo.pdf) and ending with dev.off() #(includes last column as average # sequence profile : use Ave=FALSE to eliminate this column ) d<-eg7(3,3);Td<-transformE(d) #when the data is perfectly equitable many plots #are identical for the different transforms plotsummary(Td) #points (even for average profile) have no error # in perfectly equitable system as they are specified by f,g, and u plotsummary(Td,fave=TRUE) #add noise to this signal data set #find the std dev of the overall signal and add normally distributed noise sdd<-sd(d,na.rm=TRUE) # that has a std. dev that is some fraction (fac) of this signal std dev #set the fraction of noise relative to the standard deviaiton of the signal fac<-1/3 #add to signal a normal distribution of noise with this std dev. d_noise<-d+rnorm(prod(dim(d)),mean=0,sd=fac*sdd) d_noise<-matrix(d_noise,nrow=nrow(d),ncol=ncol(d)) rownames(d_noise)<-rownames(d); colnames(d_noise)<-colnames(d) Td_noise<-transformE(d_noise) #transform the noisy data #shows how the transform looks compared to the original data plotsummary(Td_noise) #shows how the data looks compared to the signal data plotsummary(Td_noise,Td) #change the label spacing on the images to fit in the yaxis numbers plotsummary(Td_noise,yline=5,yma=8) plotsummary(Td_noise,Td,yline=5,yma=10, fave=TRUE, row_unit="Day Number", col_unit="Year", z_unit="Temperature (C)",inc=1) # plot averages of the data /signal and #compare to averages with error due to equitable system # 45x30 data set of 3 sets of random numbers coupled together #in an equitable system d<-eg8(3,3) Td<-transformE(d,Ave=TRUE) #data set entirely equitable but rows and column values have random distribution plotsummary(Td_noise=Td,fave=TRUE) # averages along rows and columns show large error but #system is entirely specified by f,g,u #no errors in knowing equitable average values as they are entiely # specified in system
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.