DR4 adaptation

[claude4334]

Describe the bug
In DR4, the length of the Least Squares solution for the continuous spectra has changed. In the BP band, there are now only 51 parameters instead of 55. The RP band remains at 55 parameters.

The XML files for the different Xp Filters (e.g. XPFilter_GaiaDr3Vega_V375wiv142.xml) define sample bases of length 55 for both XP bands, which creates a dimension mismatch between the design matrix and the length of the coefficients vector in the Multi Synthetic Photometry Generator.

To Reproduce
Execute the following python script that creates a dataframe from one DR4 source (extracted from a GBin) and calls generate()
gaiaxpy_dr4.py.txt

Expected behavior
The following exceptions is raised:
ValueError: Coefficients length doesn't match the design matrix dimension. Please make sure you're using the correct input files and configuration.

See full log attached
gaiaxpy_dr4.py.log.txt

Screenshots
n/a

Desktop (please complete the following information):

• OS: Ubuntu
• Browser Brave
• Version n/a

Additional context
n/a

[fdeangeli]

This is no bug. The public release of GaiaXPy is required to work with public data, i.e. DR3 data.

An internal release of GaiaXPy, updated to work with Cycle 04 data is being prepared as requested on the 21st of November verbally and with an issue (#112) raised on the 26th of November.

[claude4334]

Sorry for the misclassification (it did not seem like a feature request, though) and I was not aware of the discussions.

Concerning issue #112, though, the description mentions the integration of GBins as input, but that is not the object of my request. The issue I mention comes after that stage, once we have loaded the data. It is a question of matrices dimensions mismatch between the data imported from the Gbin and the PhotometricSystem model.

[yazun]

Hello,
Any news on the timeline for DR4 data compatibility?