Re: Transit Fitting plus referencing software in paper
Posted by karenacollins on Feb 14, 2015; 10:32am
URL: http://astroimagej.170.s1.nabble.com/Transit-Fitting-plus-referencing-software-in-paper-tp199p249.html
Hi David,
This is an important question for data you are planning to publish. At this time, AIJ only finds the best fit values for the model parameters. It does not find error for each of those model parameters. That is best done by a technique called Markov Chain Monte Carlo (MCMC). Wikipedia has an article on MCMC if you are not familiar with the concept. I plan to add that capability to AIJ down the road, but it will not happen for about another year.
In the mean time, there is a free web-based tool (EXOFAST) that will provide that capability using input from AIJ. The website is here:
http://astroutils.astronomy.ohio-state.edu/exofast/exofast.shtml
Instructions for the use of that tool are provided by the authors at the website.
First you would want to find the best fit model in AIJ, primarily to determine which detrending parameters you want to include in the final fit, and for partial or otherwise tricky light curves, you may need to include some model parameter 'Priors' on the webpage form that you can read off of the AIJ fit panel (i.e. the best fit model parameters shown in AIJ).
Using Multi-plot->File->Save data subset, save the following columns in the order indicated:
BJD_TDB (time)
rel_flux_T1_n (normalized relative flux - NOT detrended)
rel_flux_err_T1_n (normalized relative flux error)
detrend param 1
detrend param 2
...
detrend param n
You should use BJD_TDB as your time base. If you don't have that format in your fits headers, or didn't extract that into the measurements table, you can great BJD_TDB in AIJ using your object coordinates, observatory geographic location,a nd the JD_UTC times in the measurements table (either J.D.-240000 or JD_UTC). See this post at the support forum to help in creating BJD_TDB:
http://astroimagej.1065399.n5.nabble.com/Add-new-astronomical-values-such-as-BJD-TDB-AIRMASS-etc-to-an-exisiting-measurements-table-td91.html
The astronomical data calculation interface has changed slightly since that post, but not so much as to invalidate those instructions.
Once you have plotted normalized rel_flux_T1 (assuming your target is T1), you need to save the plotted values back to the measurements table by clicking the 'New Col' icon (
) on the left
side of the associated plot row.
Then you can use Multi-plot->File->Save data subset to save the data listed above to a text file that you can open in a text editor and then copy and paste into the photometry side of the EXOFAST webpage interface.
Karen
URL: http://astroimagej.170.s1.nabble.com/Transit-Fitting-plus-referencing-software-in-paper-tp199p249.html
Hi David,
This is an important question for data you are planning to publish. At this time, AIJ only finds the best fit values for the model parameters. It does not find error for each of those model parameters. That is best done by a technique called Markov Chain Monte Carlo (MCMC). Wikipedia has an article on MCMC if you are not familiar with the concept. I plan to add that capability to AIJ down the road, but it will not happen for about another year.
In the mean time, there is a free web-based tool (EXOFAST) that will provide that capability using input from AIJ. The website is here:
http://astroutils.astronomy.ohio-state.edu/exofast/exofast.shtml
Instructions for the use of that tool are provided by the authors at the website.
First you would want to find the best fit model in AIJ, primarily to determine which detrending parameters you want to include in the final fit, and for partial or otherwise tricky light curves, you may need to include some model parameter 'Priors' on the webpage form that you can read off of the AIJ fit panel (i.e. the best fit model parameters shown in AIJ).
Using Multi-plot->File->Save data subset, save the following columns in the order indicated:
BJD_TDB (time)
rel_flux_T1_n (normalized relative flux - NOT detrended)
rel_flux_err_T1_n (normalized relative flux error)
detrend param 1
detrend param 2
...
detrend param n
You should use BJD_TDB as your time base. If you don't have that format in your fits headers, or didn't extract that into the measurements table, you can great BJD_TDB in AIJ using your object coordinates, observatory geographic location,a nd the JD_UTC times in the measurements table (either J.D.-240000 or JD_UTC). See this post at the support forum to help in creating BJD_TDB:
http://astroimagej.1065399.n5.nabble.com/Add-new-astronomical-values-such-as-BJD-TDB-AIRMASS-etc-to-an-exisiting-measurements-table-td91.html
The astronomical data calculation interface has changed slightly since that post, but not so much as to invalidate those instructions.
Once you have plotted normalized rel_flux_T1 (assuming your target is T1), you need to save the plotted values back to the measurements table by clicking the 'New Col' icon (
) on the left
side of the associated plot row.Then you can use Multi-plot->File->Save data subset to save the data listed above to a text file that you can open in a text editor and then copy and paste into the photometry side of the EXOFAST webpage interface.
Karen
| Free forum by Nabble | Edit this page |