This is just a suggestion, but when I was messing around with Little MALLET wrapper in my Binder notebooks, I noticed that .import_data() and .train_topic_model() always report "Complete" even when Java is not installed or the MALLET path is wrong. So I was wondering if subprocess might be better than os.system for simple error reporting. However, if you capture the output with subprocess, then you can't see the topic model training on the command line...
This is just a suggestion, but when I was messing around with Little MALLET wrapper in my Binder notebooks, I noticed that
.import_data()
and.train_topic_model()
always report "Complete" even when Java is not installed or the MALLET path is wrong. So I was wondering if subprocess might be better than os.system for simple error reporting. However, if you capture the output with subprocess, then you can't see the topic model training on the command line...