Just like calibrationMarginStats
, gives stats about the calibration process:
differences between totals after/before calibration and margins. Totals for categorical
variables are displayed in percentages. The last column, named "difference", shows
the difference (in percentage points) between initial estimates and margins (if colCalibratedWeights is NULL)
or between calibrated estimates and margins (if colCalibratedWeights is not NULL).
Output is a dataframe, which might be more convenient to export than a list
(e.g. for integration into reports).
marginStats(
data,
marginMatrix,
pct = FALSE,
popTotal = NULL,
colWeights,
colCalibratedWeights = NULL,
calibThreshold = 1
)
Dataframe containing stats on weights and margins
dataframe containing the survey data
matrix of margins
Set this to true if margins for categorical variables are written in percentages
total of population, useful if margins are entered in relative value
name of weights column in the dataframe
name of calibrated weights column in the dataframe (if applicable)
If difference between calibration estimate and margin differ more than this parameter, calibration is considered to have failed
calibrationMarginStats