Those are the scripts used for the generation and plotting of the data presented in ``Magic State Injection with Erasure Qubits'' (arXiv:2504:02935).

Since some scripts are using the code published by Gidney in the "Cleaner magic states with hook injection" paper, first run step0_clone_Gidney_repo to clone its repo into the gidney_code folder.

Figure 3a & 7:

• The data was collected using expand_script.py.
• Collected data in out/expand_raw,*,2025-01-30,batch_size=500.csv.
• Plotting tools in expand_plot.ipynb.

Figure 3b:

• Data generation tools in end2end_script.py.
• Data in out/pareto_raw,*,2024-12-15,batch_size=500.csv.
• Plotting tools in plot_pareto.ipynb.

Figure 4 & 12:

• Data in out/2024-12-21-erasuremagic_1e6runs_sj for distance 3.
• Data in out/2025-07-17-erasuremagic_5e5runs_d=5 for distance 5.
• Data in out/2025-07-17-erasuremagic_5e5runs_d=7 for distance 7.
• Plotting tools in Erasuremagic.ipynb.

Figure 5:

• To decide which qubits to declare as erasure qubits, the notebook Combination_searching.ipynb was used.
• To simulate and plot the result, the notebook cultivation.ipynb was used.

Figure 6:

• The qpic code for the figure is given in qpic folder.

Figure 2 & 8 & 9:

• Data in out/2024-12-24-compare_injections_5e6runs_2erasures and out/2025-07-13-compare_injections_25e5runs_1erasure.
• Plotting tools in compare_injections.ipynb & compare_injections-vary_distance.ipynb.

Figure 10:

• To simulate and plot the result, the notebook error_counter.ipynb was used.