error estimates with montecarlo

[Letizia Stanghellini]

Hi Christophe,

I hope everything is good with you.  I am running your new montecarlo notebook on some data with errors. As far as I understand, your program in https://github.com/Morisset/PyNeb_devel/blob/1.1.19/docs/Notebooks/MonteCarlo.ipynb produced additional data within a default error of 0.5 from the original data. I would like instead to use my own measured errors, writing  adata file such as:

%%writefile observations1.dat
Target 1
H1r_4341A 48.58780784858967  0.7604427588774331
H1r_4861A 100                0.00
O2_3726A 47.77835591490365   0.7604427588774331
O2_3729A 68.8240974164782    0.9876793793287749

etc

where the errors are in the the third column and they are relative to Hbeta. I read them with

obs = pn.Observation(obsFile='observations1.dat', fileFormat='lines_in_rows', errIsRelative=True)

Is this correct? Would the montecarlo propagation then produce addiitonal observations including errors whithin the errors given?

Many thanks! Sorry if all of this is trivial!

Best wishes,

Letizia

[Christophe Morisset]

Hi Letizia,

The different format for the observation files are described in the dedicated notebook, especially cell 21 for your case:

https://github.com/Morisset/PyNeb_devel/blob/master/docs/Notebooks/PyNeb_manual_6.ipynb

If you put: LINE     TT   err

as first row of your file and read it using (adapt the values of the keywords to your case):

obs = pn.Observation('obs_test.dat', fileFormat='lines_in_rows_err_cols', corrected=False, errIsRelative=False)

you can add MonteCarlo values using:

obs.addMonteCarloObs(N=100)

This will create 100 fake observations, named TT_MC_1 to TT_MC_100, taking into account the error of each line

You can have more than one observation in the original file, just be sure thier names are different.

To answer you r other question in this same email, you can also have the emissin lines in cols and the objects in row, have a look at the cell 19 in the same Notebook.

Hope it helps,

Saludos,

Christophe