Hello Group 11!
My name is Mai Le, I'm assigned to do a peer review for your group project this week. Overall, I like your topic, interesting dataset, and good usage of a prediction model, you have a good repo structure, your files are well written and easy to understand.
I have a few suggestions below (already reviewed TA's feedbacks to avoid overlapping issues):
README: there seems to be missing a section related to Makefile, which should have been included as part of the last milestone
I have followed the instruction, install and activate conda environment before running run_analysis.sh, and got these issues related to missing libraries (shortening the error messages here):
Error in library(tidyverse) : there is no package called ‘tidyverse’
Error in loadNamespace(name) : there is no package called ‘rmarkdown’
seismophobia_report.Rmd:
Figure 2: color scale is hard to read
Figure 5: table figure's style looks a bit out of place, do you consider using kable for better visualization?
Low F1 score? Have you tried RandomForestClassifier? And probably address the class imbalance by using class_weight? (probably overlapping with TA's feedback)
I know you probably have already thought about all of those, we have just too little time for things we want to do in this project. Hopefully by giving feedback at the beginning of the week, you will have more time to work on these.
Hello Group 11! My name is Mai Le, I'm assigned to do a peer review for your group project this week. Overall, I like your topic, interesting dataset, and good usage of a prediction model, you have a good repo structure, your files are well written and easy to understand.
I have a few suggestions below (already reviewed TA's feedbacks to avoid overlapping issues):
README: there seems to be missing a section related to Makefile, which should have been included as part of the last milestone
I have followed the instruction, install and activate conda environment before running
run_analysis.sh, and got these issues related to missing libraries (shortening the error messages here):seismophobia_report.Rmd:
I know you probably have already thought about all of those, we have just too little time for things we want to do in this project. Hopefully by giving feedback at the beginning of the week, you will have more time to work on these.
All the best.