Next steps#

No plots, just exposure times#

You can also calculate the exposure time required to obtain a given S/N using the signal_to_noise_to_exp_time function. For example - how many seconds must one expose ARCES on a V=12 mag M0V star to get a S/N of 30 at the wavelength of H-alpha:

from arcesetc import signal_to_noise_to_exp_time
import astropy.units as u

sptype = 'M0V'
wavelength = 6562 * u.Angstrom
signal_to_noise = 30
V = 12
print(signal_to_noise_to_exp_time(sptype, wavelength, V, signal_to_noise))

This returns 642.11444 s, a Quantity object containing the required exposure time.

Available spectral types#

You can see which spectral types are available with the available_sptypes function.

Note

At present, the best coverage is for mid-F through mid-M type main sequence stars, with some M giants.

Here’s a color-magnitude diagram of the stars presently available in arcesetc:

../_images/cmd.png

How it works#

We estimate the count rates for stars as a function of wavelength by fitting 15th-order polynomials to each spectral order of real observations of a star of each spectral type. These polynomial coefficients and some wavelength metadata are stored in an HDF5 archive for compactness and ease of reconstruction. Then upon calling arcesetc, the archive is opened and the spectral order closest to the wavelength of interest is reconstructed from the polynomial coefficients, for a star of the closest available spectral type to the one requested.

Warning

arcesetc doesn’t know anything about saturation.

Run the tests#

If you’re contributing to arcesetc, you can check that your updates don’t break the API by running the tests like this in the source directory:

tox -e test

If the tests pass, you’re ready to submit a pull request!