StackOverflow2013

Note that there are some explanatory texts on larger screens.

plurals

PO
text
Body
copied!<p>As an analytical technique acf/pacf/ccf are used to identify periodicity in time-dependent signals, hence the correlogram, the graphical display of acf or pacf, displays self-similarity in a signal as a function of different lags. (So for instance, if you see values on the y axis peak at a lag of 12, and if your date is in months, that's evidence of annual periodicity.) </p> <p>To calculate and plot 'similarity' versus lag, if you don't want to roll your own, i am not aware of a native Numpy/Scipy option; I also couldn't find one in the 'time series' scikit (one of the libraries in the Scipy 'Scikits', domain-specific modules not included in the standard Scipy distribution) but it's worth checking again. The other option is to install Python bindings to R (RPy2, available on SourceForge) which will allow you to access the relevant R functions, including '<strong>acf</strong>' which will calculate and plot the correlogram just by passing in your time series and calling the function.</p> <p>On the other hand, if you want to identify continuous (unbroken) streams of a given type in your signal, then "<strong>run-length encoding</strong>" is probably what you want: </p> <pre><code>import numpy as NP signal = NP.array([3,3,3,3,3,3,3,3,3,0,0,0,0,0,0,0,0,0,0,7,7,7,7,7,4,4,1,1,1,1,1,1,1]) px, = NP.where(NP.ediff1d(signal) != 0) px = NP.r_[(0, px+1, [len(signal)])] # collect the run-lengths for each unique item in the signal rx = [ (m, n, signal[m]) for (m, n) in zip(px[:-1], px[1:]) ] # returns [(0, 9, 3), (9, 19, 0), (19, 24, 7), (24, 26, 4), (26, 33, 1)] # so w/r/t first 3-tuple: '3' occurs continuously in the half-open interval 0 and 9, and so forth </code></pre>

Querying!

Guidance

An individual column

Larger individual text columns get their own page to allow for proper reading.

SQuiL has stopped working due to an internal error.

If you are curious you may find further information in the browser console, which is accessible through the devtools (F12).

Reload