Where is `plot_chi`

[reducedpc]

I am following this tutorial page on "Solving the linearized Eliashberg equation".

In the step 1. Construct the density- and magnetic-susceptibilties in RPA , the cell number [8] says,

plot_chi(chi_m_wk, chi_d_wk).

But where does the function plot_chi(arg0, arg1) exist? Or is there a dedicated page for all TRIQS plotting functions?

[Thoemi09]

Hi @reducedpc,

I think this is a special function just for the tutorials. You can find it here: https://github.com/TRIQS/tprf/blob/3.3.x/doc/user_guide/Solving%20the%20linearized%20Eliashberg%20equation%20in%20the%20random%20phase%20approximation%20limit.ipynb

I'm not sure why it doesn't show up on the documentation page. Thanks for pointing it out.

[reducedpc]

Yes, I just discovered that as well. However, it's throwing errors:

-```

AttributeError Traceback (most recent call last)
Cell In[8], line 1
----> 1 plot_chi(chi_m_wk, chi_d_wk)

Cell In[1], line 47, in plot_chi(chi_m_wk, chi_d_wk)
38 path = [(r'$\Gamma$', 2np.pinp.array([0.0, 0.0, 0.0])),
39 ('X', 2np.pinp.array([0.5, 0.0, 0.0])),
40 ('M', 2np.pinp.array([0.5, 0.5, 0.0])),
41 (r'$\Gamma$', 2np.pinp.array([0.0, 0.0, 0.0])),
42 ]
45 ax_bs = plt.subplot(111)
---> 47 ax_bs.bsplot(chi_m_wk[(Idx(0), slice(None))], path)
48 ax_bs.bsplot(chi_d_wk[(Idx(0), slice(None))], path)
50 ax_bs.set_ylabel(r'$\chi(i\nu_n=0, \mathbf{k})$', rotation=0, ha='right')

File ~/miniconda3/envs/jupyter_env/lib/python3.10/site-packages/triqs_tprf/plotting_tools.py:76, in (self, obj, path, *opt_list, **opt_dict)
72 top.spines['left'].set_bounds(lower_limit, upper_limit)
73 top.set_yticks([lower_limit, upper_limit])
75 mpl.axes.Axes.bsplot = lambda self, obj, path, *opt_list, **opt_dict :
---> 76 __bsplot_impl(self, obj, path, *opt_list, **opt_dict)
78 # ========== DOS ==========
80 def dosplot(obj, *opt_list, **opt_dict):

File ~/miniconda3/envs/jupyter_env/lib/python3.10/site-packages/triqs_tprf/plotting_tools.py:50, in __bsplot_impl(top, obj, path, *opt_list, **opt_dict)
47 hs_labels, hs_k = list(zip(*path))
49 k_paths = list(zip(hs_k, hs_k[1:]))
---> 50 k_vecs, k_plot, K_plot = k_space_path(k_paths, bz=obj.mesh.domain)
51 kx, ky, kz = k_vecs.T
53 # -- If top is plt do now access the currently used Axes object

AttributeError: 'triqs.gf.meshes.MeshBrZone' object has no attribute 'domain'

[Thoemi09]

What is your TRIQS and TPRF version?

[reducedpc]

Both are 3.3.1.

[Thoemi09]

It seems that bsplot and therefore also plot_chi don't work anymore with newer TRIQS versions. It tries to use mesh.domain which has been removed in v3.2.

I can try to fix it but it might be good to get an opinion from @HugoStrand or @Wentzell.

Thanks a lot @reducedpc for finding that problem.

[reducedpc]

Thanks @Thoemi09. Alright, I'll wait for @HugoStrand or @Wentzell's response then. If you let me know where I can view the structure of bsplot, that would be helpful as well.

[Thoemi09]

Sure, no problem. You can find it here.

If you are looking for specific functions/classes/variable names/etc., I can recommend trying GitHub's search option in the top right corner. It lets you search the specific repository/organization/etc. that you are currently viewing. For example, typing bsplot should lead you right to the link above.

Best,
Thomas