Public Data Access Overview / Discussion Forum

Hi,
I was using the catalog containing SDSS ugriz colours for TNG300-1, redshift 0. However, I could not see any bimodality in the (u-r) colour distribution(which is visible in observational data). I am not sure if I am doing any mistake.
Here is the code snippet:

hf = h5py.File('Subhalo_StellarPhot_p07c_cf00dust_res_conv_ns1_rad30pkpc_099.hdf5', 'r')
selected_key = 'Subhalo_StellarPhot_p07c_cf00dust_res_conv_ns1_rad30pkpc'
urcolour = hf[selected_key][:,0,0] - hf[selected_key][:,2,0]

Is there any other criterion ?

Perhaps you would like to select a particular stellar mass range?

At first, I was considering the whole mass range. Later I found a paper on TNG300 by you (First results from the IllustrisTNG simulations: the galaxy colour bimodality ) and took 9<log(stellar mass) < 12 . Then I got the bimodal distribution of (u-r) colour.
However, the mass distribution is not of a known particular shape ( e.g. gaussian or lognormal).
Should we consider this particular mass range to get the observed colour distribution?
Thanks!

If your goal is a comparison with data, you would want to choose the same mass range (and any other characteristics) of the data.