Optimizer issue

[kinsey40]

We look at performing iterations of a Gaussian distribution of data across each data points error bars. The optimizer is used to examine the best period of the L-S model. A histogram plot is created of the no. of iterations against the value for the optimized period. We observe a large spike at a particular value (1763 days) for no apparent reason.

The optimizer is compared to a 'MAX' method whereby the best period is selected through examination of the max power of the L-S model fit.

The results highlight that often, over 50% of the optimizer chosen periods fall onto the same exact value as another iteration (up to ~ 10 sig figs) however, for the MAX method only a small percentage exhibit such behaviour (heavily influenced by oversampling fact size).

We see no reason for such a difference to arise and can only conclude that the optimizer for whatever reason does not seem to work for this particular data.
Note that the optimizer does appear to work for other objects that we have examined.

Believe code is implemented correctly, although mistakes/mis-understandings may be present.

Code located in Word Doc; Data in Text document (Download to same file location to run properly)

Thanks.

Gatspy_Issue.docx

Gatspy_issue_Data.txt