Problem on using CAFE_fig to visualise the output of CAFE5

[YuzhouGong]

Hello Folks,
I realised this issue might be similar to #69 (comment), but I think there are unsolved questions in the closed issues 69.
I found that the .cafe file genereted by CAFE5 is different from the .cafe file generated by CAFE4 (4.2.1), like @Xacfran has mentioned in issues 69, the Viterby-P-values column in CAFE4' result is not presented by CAFE5, which might casue wrong outputs of CAFE_fig (all nodes show +0/-0 expanded and contracted gene family, when not using --count_all_expansions option in CAFE_fig). The result of my data, looks like this:

I used command: python3 CAFE_fig.py Base_report.cafe --dump test/, and some warning messages have jumped out:
"Parsing CAFE report... done!
Writing test/summary.pdf
qt.qpa.xcb: failed to initialize XRandr
qt.qpa.xcb: X server does not support XInput 2
qt.qpa.xcb: XKeyboard extension not present on the X server"
, and generated the above figure.
If I added --count_all_expansions in my command, the tree nodes will show the numbers of all expanded and contracted gene families, but still a condensed tree shape, like this:

I know there is a visualization tools designed specifically for CAFE5 called CAFE_plotter, but it could only show numbers of all expanded and contracted gene families on the tree (this was confirmed by its author).
As in my case, I am only interested in presenting the number of gene families which have gone through change with statistical significance, instead of showing the big number of all expanded/contracted gene families, is there a easy way by using CAFE5, like omitting count_all_expansions in CAFE_fig?
(By the way, I know how to caculate the number of significantly expanded/contracted gene families of tree nodes using awk and grep command from CAFE5 result files, for example *_family_results.txt and *_change.tab).