I am checking the association of cold with hospital admissions. I can easily check the association of mean daily temperature (as an exposure variable) on admissions as it'a linear. However, i want to check the coldspell variable association ( a categorical variable with values 1= yes (coldspell) 2= no.. So i want to check the effect of cold spells on outcome. Here are the sample plots that i want to get (i.e. ¨Figure 2 and 3).
https://www.nature.com/articles/srep38380?proof=true%3Ca+href%3D#Sec18
I am using following in crosspred for linear plots. What changes should i make?
predglm
You could try the following if you code cold spell as 1 and non-cold spell as 0.
pred.all
This paper (pg. 4) can help you. Here, Z represents cold spell exposure, which was defined as a categorical variable (2 = 2008 cold spell days, 1 = cold spell days in 2006, 2007, 2009 and 2010, and 0 = non-cold spell days).
https://ehjournal.biomedcentral.com/track/pdf/10.1186/1476-069X-13-60.pdf
- Badges
- Science topic
- Similar topics
- Geoscience
- Atmospheric Sciences
- Climate Science
- Climate Change