findHighestChis {beadarray} | R Documentation |
Function for finding which bead types on an array are least randomly positioned.
findHighestChis(BLData, array, limit = 14)
BLData |
an BeadLevelList object containing bead level data |
array |
numeric value for which array we want to use |
limit |
numeric value which determines the threshold value for the chi-statistic. |
The $chi^2$ statistic is defined as:
begin{center}
$chi^2 = sum^k_{i=1} (O_i - E_i)^2 / E_i$ end{center}
where O and E are vectors of length 8 defining the observed and expected number of beads in each section respectively. This provides a means of systematically testing the distribution of every bead type on a given array and we can record each bead type with a sufficiently high value of $chi^2$.
If the value for a particular bead type is found to be greater than 'limit' then we add the ProbeID for that bead type to a list which we later return.
Default is to use 14 as the limit as this is the 95th quantile for the chi-square distribution in question
numeric vector giving ProbeIDs for all bead types on the array which give a chi-statistic > 'limit'
Mark Dunning
data(BLData) ##Default setting is to find all bead types with a chi-statistic greater than 14 findHighestChis(BLData, 1) ##Now find all beads with chi-statistic for their distribution greater than 16 findHighestChis(BLData, 1, limit=16) ##ProbeIDs returned by the function can be examined in more detail plotBeadLocations(BLData, ProbeID=278, array=1) probeDiagnostics(BLData, 278, 1)