Data Transformation in predmicror

Consistent Use of Natural Logarithm

In previous versions of predmicror, there was an inconsistency in the required data transformation for different models. Some models required the data to be in natural logarithm (ln) scale, while others required it to be in base-10 logarithm (log10) scale. This could lead to confusion and errors.

To address this issue, all models in predmicror have been harmonized to use the natural logarithm (ln) for the response variable Y(t).

This means that users should always provide the microbial concentration in ln scale.

Converting from log10 to ln

If your data is in log10 scale, you can easily convert it to ln scale using the following formula:

ln(N) = log(10) * log10(N)

Here is an example of how to convert a column in a data frame:

# Create a sample data frame with log10 data
my_data <- data.frame(
  Time = c(0, 1, 2, 3),
  log10N = c(2, 2.5, 3, 3.5)
)

# Convert the log10N column to lnN
my_data$lnN <- log(10) * my_data$log10N

# Print the updated data frame
print(my_data)
#>   Time log10N      lnN
#> 1    0    2.0 4.605170
#> 2    1    2.5 5.756463
#> 3    2    3.0 6.907755
#> 4    3    3.5 8.059048

By ensuring that all models use a consistent ln scale, predmicror is now more user-friendly and less prone to errors.