Michael Hippke

``` if not isinstance(flux, np.ndarray): print("Provided flux array is not a numpy array. This may slow down nested sampling considerably. Please consider converting the array: flux = nparray(flux)") ```

In the [code](https://github.com/hippke/Pandora/blob/b7da6f78dc66a8969e4d343e60c7692268556608/pandoramoon/pandora.py#L360), a day is simply defined to have `60*60*24` seconds. The `params.cadences_per_day` are then just an `int` based on that number. If you have a hard requirement to...

There is currently no way to directly limit transit duration, neither max nor min. Their values are calculated based on physical limits ([link to code line](https://github.com/hippke/tls/blob/4ef0f141d842377dd2541906d84c2e4350027c27/transitleastsquares/grid.py#L35)). Ad hoc, you could...

Do you have a good example to test? Code + data

Ah, I see. At the time I tested TLS, I used a lot of Kepler and early TESS data with only minor gaps. So that's a use case I had...

Sure, but what's it good for? The user can already request a grid with any [from, to] in any log-spacing they want?

Yes, you should be able to use any floating point value for `oversampling_factor`, e.g. 0.5

Looks like a bug. Can you share the code to reproduce? I will look into it and make a bugfix if possible.

Fixed in version [1.0.26](https://github.com/hippke/tls/releases/tag/1.0.26), which is also available via pip. `oversampling_factor` now accepts any floating point value.

Sorry you are having trouble. Looking into it right now. The runtime of your script is ~1 minute on my Core-i7 7700k using 7 threads (and considerable background load). Your...