search

BrunaLab / bdffpRain

R package for rain gauge data from the Biological Dynamics of Forest Fragments Project.
MIT License
0 stars 0 forks source link

Missing values possibly more common in the dry season #9

Open Aariq opened 4 years ago

Aariq commented 4 years ago

NA's are not evenly distributed throughout the year. They seem to be more common in the dry season, indicating to me that at least some of the NAs might actually be zeroes. Is there another explanation for this? Maybe when fieldwork was done or observations were made? distribution of NAs

embruna commented 4 years ago

I would bet that these NA in dry season (approx may-dec) are actual zeros.

Might be worth (mentioned on another issue) looking at this in different years and for less used vs. more used camps. Also some minor edits: floresta->florestaL (missing L at end) 41dia->km41 Cabo->CaboFrio Porto->PortoAlegre 37->km37 (I figured the last two were because of the 2 word names, so whatever you think is easiest. PA and CF are options i guess, but CF is also the abbreviation for Continuous Forest...)

embruna commented 4 years ago

Is the y axis here "Proportion of missing values that are NA?"

Aariq commented 4 years ago

Is the y axis here "Proportion of missing values that are NA?"

No it's proportion of dates that are in the spreadsheets that have no entry for precipitation.

Aariq commented 4 years ago

https://github.com/BrunaLab/Heliconia-Drought/issues/1#issuecomment-655026589

Aariq commented 4 years ago

This was due to a mistake on my part. I updated code that was unexpectedly deleting zeroes and now the distribution of missing values is more even. Still looks like there might be some seasonality to it, which may be worth noting in any description of this dataset. I'll leave this issue open for now. distribution of NAs