search

tnoulas / UrbanEpidemicSim

5 stars 2 forks source link

Population moving at run_simulation() #2

Open kpelechrinis opened 4 years ago

kpelechrinis commented 4 years ago

I was getting an error at the following line:

moving_pop = random.sample(venue1_population_set, moving_population_size)

and the reason is that moving_population_size is defined as:

moving_population_size = 1.0 / self.places[venue1].get_total_movements()

random.sample() expects an integer as its second argument, so I assume that since each entry at the df_transition_snap is one transition, we can have:

moving_pop = random.sample(venue1_population_set, 1)

Also at the population exchange, the following line was giving me a Typerror (TypeError: unhashable type: 'list'):

new_venue2_pop = venue2_population_set.add(moving_pop)

and I changed it to

new_venue2_pop = venue2_population_set.union(moving_pop)

and works now.

Also I think that the following two lines at the end of this function should be one level of indentation to the left:

epoch+=1 date1 = date2

tnoulas commented 4 years ago

Strange, did you pull an older version? I had the designated line like that: moving_population_size = int(1.0 / self.places[venue1].get_total_movements() )

Again for the second bit, try to pull the newest version :)

I am not sure when you forked!

kpelechrinis commented 4 years ago

Hmm is it the https://github.com/tnoulas/UrbanEpidemicSim/blob/master/UrbanEpidemicsSim.ipynb ? Because if it is, I still see the line for the moving_population_size without the integer casting.

tnoulas commented 4 years ago

No this is the notebook version which is not up to date I think! Try to use python PlaceNetSim_v2.py as I noted in my last email. I will update the notebooks later. I am sorry for the confusion.

tnoulas commented 4 years ago

I am sorry about it. I even realised that I have two notebooks with slightly similar name ... That's what happens when you do things too fast I guess...

tnoulas commented 4 years ago

Reopening this as it happens to be the same line of code you mentioned that is buggy: moving_population_size = int(1.0 / self.places[venue1].get_total_movements() ) has been actually always reason.

for the time being I am setting each transition to just move one person: moving_population_size = 1

will try to think about it tomorrow with clear head :)

andylamp commented 4 years ago

@tnoulas can take a look at it as well... maybe I can be of help.