must download both Cleaning the Crew.ipynb, CleaningFunctions.py to work
and I have made multithread proccessing avaliable to reduce the running time from 15 mins to ~30 sec!!! SO NICE!!!
#dropping the columns that was not needed
#filling empty list with []
#changing a string of list into list
#changing the dict in the list into ID only
# getting all unique values
# adding the extra columns for the most related crew (improved on timing!!!)
#changing the time into numerical value (improved on timing!!!)
# save to save.csv
#dropping the extra columns that was already extracted
#creating df df_pred to do prediction for those with non-nan values > 2000
#cause knn dosent accept a string heading, so changng the heading
# KNN avaliable
#adding back the columns for df_pred
#merging df with df_pred