search

tyjo / clv

Compositional Lotka-Volterra
MIT License
15 stars 7 forks source link

CompositionalLotkaVolterra(P, T, U, pseudo_count=1e-5) #2

Closed gaofengjie-2019 closed 3 years ago

gaofengjie-2019 commented 3 years ago

Dr.Tyler, Sorry to bother you. Here I found a error when I run the function: CompositionalLotkaVolterra(P, T, U, pseudo_count=1e-5) That is: clv = CompositionalLotkaVolterra(P, T, U, pseudo_count=1e-5) Error: no valid denominator found Traceback (most recent call last):

File "", line 1, in clv = CompositionalLotkaVolterra(P, T, U, pseudo_count=1e-5)

File "D:\pydata\clv\compositional_lotka_volterra.py", line 104, in init self.denom = choose_denom(P)

File "D:\pydata\clv\compositional_lotka_volterra.py", line 62, in choose_denom exit(1)

NameError: name 'exit' is not defined

It indicates no valid denominator found. Maybe there are some problems in data from P or Y?The concentrations of each taxa in Y are the relative abundance, is it OK? Fengjie

tyjo commented 3 years ago

What dataset are you running this on? What value are you using for the pseudo count?

Since cLV works under the additive log-ratio transformation, the code tries to make a sensible choice for the denomination. Specifically, it looks for a taxon present in all samples where the variance of the log change over time is small.

Since the denominator choice procedure requires taking a log, I suspect there might be zeros in the relative abundances (P).

gaofengjie-2019 commented 3 years ago

What dataset are you running this on? What value are you using for the pseudo count?

Since cLV works under the additive log-ratio transformation, the code tries to make a sensible choice for the denomination. Specifically, it looks for a taxon present in all samples where the variance of the log change over time is small.

Since the denominator choice procedure requires taking a log, I suspect there might be zeros in the relative abundances (P).

Thanks a lot for your reply. I have solved this problem.