Prepare Data for Outcome Stability Analysis

This function prepares data for use with the StabCalc function. It calculates relative, absolute and adjusted coefficients of yield variation for unique temporal sequences of at least three years.

PrepareStabData(Data, OutCodes = 101)

Arguments

Data

a data.frame or data.table of ERA data with column names standardized using the StandColNames function.

OutCodes

A vector of outcome codes to consider in the analysis. Default = 101 (Crop Yield).

Value

a data.table including fields of the data supplied, subset to unique temporal sequences with the following columns appended:

  • UID integer, a unique identifier for a pairing of control and experimental treatments (CT and ET) and their outcome that is used to match observations across time.

  • UID.C integer, a unique identifier for a CT that is used to match observations across time.

  • UID.T integer, a unique identifier for an ET that is used to match observations across time.

  • yieldcont numeric, the mean yield of the CT.

  • minyieldcont numeric, the minimum yield of the CT.

  • yieldexp numeric, the mean yield of the ET.

  • sdcont numeric, the standard deviation of the CT.

  • varcont numeric, the variance of the CT.

  • sdexp numeric, the standard deviation of the ET.

  • varexp numeric, the variance of the ET.

  • cvcont numeric, the coefficient of variation for the CT as sdcont/yieldcont.

  • acvcont numeric, the adjusted coefficient of variation for the CT.

  • cvexp numeric, the coefficient of variation for the ET as sdexp/yieldexp.

  • acvexp numeric, the adjusted coefficient of variation for the ET.

  • yieldratio numeric, the ratio of ET:CT yields as yieldexp/yieldcont.

  • sdratio numeric, the absolute variance ratio of ET:CT yields as sdexp/sdcont.

  • cvratio numeric, the relative variance ratio of ET:CT yields as cvexp/cvcont.

  • acvratio numeric, the adjusted relative variance ratio of ET:CT yields as acvexp/acvcont.

  • Rep integer, the number of replicates for the temporal sequence.

  • nryears integer, the number of yield observations in a temporal sequence.

  • cluster_cont integer, within a study (Code field) temporal sequences that share the same value have the same CT across multiple ETs.

  • UID.C.Len integer, the number of temporal sequences that have the same cluster_cont value within a study.

  • cluster_exp integer, within a study (Code field) temporal sequences that share the same value have the same ET across multiple CTs.

  • UID.T.Len integer, the number of temporal sequences that have the same cluster_xp value within a study.

  • UID1 integer, a unique identifier for each UID within outcomes (UID is unique across all the data).

  • N.Obs integer, the number of temporal sequences for a UID (this should equal one).

  • N.Obs.Study integer, number of temporal sequences contribude by a study (a unique Code field value) for a Practice x Outcome combination.

  • Weight.Study numeric, a weighting value calculated as (Rep^2/(2xRep))/N.Obs.Study.

  • Weight numeric, a weighting value calculated as (Rep^2/(2xRep))/N.Obs.

Details

The data supplied needs to have Practice and Outcome fields as generated using the ERAg::StandColNames function.

The units of measurement (the Unit field) for the supplied data should contain values of only "Mg/ha", although there is limited functionality to harmonize other units ("kg/ha", "kg/acre", "Kg/fed", "Mg/fed", and "g/m2") to "Mg/ha".

The Adjusted Coefficient of Variation (acv) is calculated using the metan::acv function.