zscores manpage

COMMAND

zscores -- Calculate standard scores of numerical data spines.

zscores [input] [> output]

-f #		Select spine # in the input for processing. The first non-**kern spine is the default.
-s		Display measured mean and standard deviations of input data only (not the resulting z-scores).
-i		Display only the input data as a single-column list of numbers.
--raw		Display the output data as a single column of numbers.
-S		Suppress the printing of the mean and standard deviation in output.
-a		Append analysis data to input contents.
-p		Prepend analysis data to input contents.
--replace		Substitute analysis data in original data position in the input data.
--full		Display full location of extracted/processed numbers in input.
-m #		Use the specified mean when calculating the z-scores.
-d #		Use the specified standard deviation when calculating the z-scores.
--sample		Use the sample standard deviation rather than the population standard deviation.

zscores converts spines of numerical data into values equivalent to the standard deviation scale by shifting the average value of the sequence to zero, and dividing by the standard deviation to generate standard scores. This allows separate sets of numbers with different ranges to be compared more directly.

To convert a set of numbers to their equivalent z-scores, the following equation is used:



By default zscores uses the population standard deviation:



If you need to use the sample standard deviation:

you can use the --sample option. The sample standard deviation is used when you have only a small number of values which probably do not match the actual mean and standard deviation if you were to have a larger set of numbers generated in the same manner (the population). Using a small subset of a population usually results in a less accurate estimate of the underlying mean and standard deviation of the entire population which the sample standard deviation tries to correct by increasing the standard deviation slightly.

Z-scores can be used to calculate Pearson correlation:

where r is called the "r-value" (i.e., the Pearson correlation) between two sequences; z_x are the z-scores for the first sequence to compare, z_y for the other sequence, and N is the length of the two sequences.

Example usages of the zscores program are avaliable on the zscores examples page.

• Wadsworth Statistics Workshop on z Scores

The compiled zscores program can be downloaded for the following platforms:
• Linux (i386 processors) compiled on 14 Jan 2008.

The source code for the program was last modified on 13 Jan 2008.