'list index out of range' Error in Data Analysis Module.

[smyconan]

I've done the process of model space generation & model selection. But errors appear in data analysis which are as the following:

Traceback (most recent call last):
  File "run_AutoCD.py", line 82, in <module>
    main()
  File "run_AutoCD.py", line 78, in main
    run_data_analysis.main(data_analysis_config)
  File "/home/AutoCD/data_analysis/run_data_analysis.py", line 35, in main
    model_space_report_df = ABC_DA.write_combined_model_space_report(finished_exp_final_population_dirs, combined_analysis_output_dir) 
  File "/home/AutoCD/data_analysis/data_analysis_ABCSMC.py", line 855, in write_combined_model_space_report
    model_space_report_df = merge_model_space_report_df_list(model_space_report_list)
  File "/home/AutoCD/data_analysis/data_analysis_ABCSMC.py", line 46, in merge_model_space_report_df_list
    model_space_report_df = df_list[0]
IndexError: list index out of range

I think that it's because the df_list is empty, but why it happens when I've already done the model space generation & model selection? By the way, I'm just following your example bash script at the bottom of your README.

I'm not familiar with the microbial community design, so maybe the problem is minor for yours. But I'm still looking forward to your reply, thank you!

[behzadk]

Hi,
It's looking for the output data from ABC SMC. Can you see these outputs in the different experiment folders?

Perhaps you need to change the DA_config yaml file so it's pointing to the correct output folder?

[smyconan]

Hi,
It's looking for the output data from ABC SMC. Can you see these outputs in the different experiment folders?

Perhaps you need to change the DA_config yaml file so it's pointing to the correct output folder?

The yaml files of configuration is like the following:

MSG two strain:

# Set absolute paths to output of ModelSpageGenerator (MSG) 
model_space_name: input_files_two_strain_0

# Output model space to the ABC input files dir
output_dir: "./ABC_input_files/"

# Paths referring to .csv files containing 
# default parameter and initial species prior ranges
default_parameters_path: './ModelSpaceGenerator/default_params/default/default_params.csv'
default_initial_species_path: './ModelSpaceGenerator/default_params/default/default_init_species.csv'

ABC two strain:

experiment_name: two_strain_0_SMC
population_modules_path: './ABC_input_files/input_files_two_strain_0/population_modules.so'
inputs_folder: './ABC_input_files/input_files_two_strain_0/input_files/'
output_folder: './output/'

DA two strain

experiment_name: two_strain_0_SMC
inputs_folder: './ABC_input_files/input_files_two_strain_0/'
output_folder: './output/'

And initially, the ABC_input_files and output folders are empty.
After the MSG process (it says No ABC config, skipping ABC..., but I think it's not a case since ABC process doesn't run yet.), the ABC_input_files folder is like the following:

.
`-- input_files_two_strain_0
    |-- adj_matricies
    |-- input_files
    |-- model.cpp
    |-- model.h
    |-- population_modules.so
    `-- py_eqs_txt_files

Then after the ABC selection process (it still says No ABC config, skipping ABC... , I'm not sure if it is OK.), the output folder is like the following:

.
|-- two_strain_0_SMC
|-- two_strain_0_SMC_0
|   |-- ABC_two_strain.yaml
|   |-- Population_0
|   |-- Population_1
|   `-- Population_2
`-- two_strain_0_SMC_0_pop_2.tar.gz

Then I run the DA process, it shows the error above.

I'm not sure if the issue comes from the difference of experiment name two_strain_0_SMC and two_strain_0_SMC_0, or the skipping ABC message, or the DA config should include the path of two_strain_0_SMC instead of the ABC input path?

Thank you very much and waiting for your reply ^ ^

[behzadk]

Ok, I see the problem. The folder two_strain_0_SMC_0 should be inside two_strain_0_SMC. I'll fix this later in the week. In the mean time, if you manually change the file structure you should be able to run the data analysis

[smyconan]

Ok, I see the problem. The folder two_strain_0_SMC_0 should be inside two_strain_0_SMC. I'll fix this later in the week. In the mean time, if you manually change the file structure you should be able to run the data analysis

Thank you behzadk! I've solve this problem by moving the folders, and can run the whole process for two strain.

But when I run the three strain process, I met gcc-9: fatal error: Killed signal terminated program cc1plus compilation terminated. error, which I think is because of the memory space? Is there any way to solve this issue? I've searched on google and found those solutions don't work.

By the way, what scale of hardware resources did you use in this experiment?

Best,

[behzadk]

What are you running when you encounter the memory error?

[smyconan]

What are you running when you encounter the memory error?

I've already solved this issue ~ It's since that the three-strain task needs more memory resources to run. I've enlarged the memory size and it works now.

Thank you very much!