Heat stress performance data

Data formatting, checking

Raw Performance Data

Read whole peformance data and start filtering:

1. keep only data from farms with temperature records
2. keep only records of closed tests (on-test and off-test data available)

[1] 86235    22

Total records read

dim(Heat_Stress_1)

[1] 175081     22

Total records from farms of interest

dim(Heat_Stress_Filter)

[1] 86235    22

Total complete records from farms of interest

dim(Heat_Stress_Drop)

[1] 82033    22

summarize the ADG data by overlapping periods in each barn

Find the timeperiods in each barn when pigs were tested

[1] 63427    16

There are a total of 881 unique barn-specific on-test/off-test combinations. But many of these overlapp substantially

Look at overlapp between those periods and create unique non-overlapping periods

There are 138 non-overlapping periods. By collapsing the on-test to the off-test periods there is minimal loss of information as the overlap is always substantial. For instance:

   Min. 1st Qu.  Median    Mean 3rd Qu.    Max. 
 0.6183  0.8077  0.8557  0.8513  0.9091  1.0000 

Now fit models to data from the non-overlapping periods to look at the sex means and differences