07.05.2015 Views

Download the documentation - True BASIC

Download the documentation - True BASIC

Download the documentation - True BASIC

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

Simple Statistics 31<br />

KurtosisP (data(), k)<br />

Percentile coefficient of kurtosis: .5 * (Q 3<br />

– Q 1<br />

) / (P 90<br />

– P 10<br />

).<br />

Exceptions:<br />

715 Can’t use KurtosisP with percentiles 10 and 90 equal.<br />

761 Need at least 99 data items for Percentiles.<br />

Moment (data(), r, mr)<br />

Given r, and dataset data(), this computes <strong>the</strong> r th moment mr about <strong>the</strong> dataset’s mean.<br />

Exceptions:<br />

716 Moment ‘r’ must be a positive integer: r<br />

MomentAbout (o, data(), r, mr)<br />

Given r, an arbitrary origin o and dataset data(), compute <strong>the</strong> r th moment mr about o<br />

as shown by <strong>the</strong> formula where n = Size(data).<br />

n<br />

Σ (data(i) – o)r<br />

i = 1<br />

————————<br />

n<br />

Exceptions:<br />

716 Moment ‘r’ must be a positive integer: r<br />

Quartiles and Percentiles<br />

These routines return <strong>the</strong> quartiles and percentiles of a dataset.<br />

Quartiles (data(), q1, q2, q3)<br />

Quartiles of a dataset data(), returning <strong>the</strong> first quartile in q1, second (median) in q2,<br />

and third in q3. Your dataset must contain at least 3 non-missing items. Quartile values<br />

will be interpolated when <strong>the</strong> quartile falls between two data items.<br />

Exceptions:<br />

762 Need at least 3 data items for Quartiles.<br />

Percentiles (data(), p())<br />

Percentiles of a dataset data() in <strong>the</strong> array r() which is automatically redimensioned to<br />

p(1:99). Thus, for example, p(95) will contain <strong>the</strong> 95% percentile value. Your dataset<br />

01/01

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!