Coming from nhanesA

R's nhanesA covers the same data. The call names map across directly.

nhanesA (R)	NHANES.jl
`nhanes("DEMO_J")`	`NHANES.download("DEMO_J")`
`nhanes("DEMO_J", translate=FALSE)`	`NHANES.download("DEMO_J"; translate=false)`
`nhanesTables("EXAM", 2005)`	`NHANES.tables(:EXAM, 2005)`
`nhanesTableVars("EXAM", "BMX_D")`	`NHANES.variables("BMX_D")`
`nhanesCodebook("DEMO_D", "RIAGENDR")`	`NHANES.codebook("DEMO_D", :RIAGENDR)`
`nhanesTranslate("BPX_D", "BPXSY1", data=df)`	`NHANES.translate!(df, "BPX_D", :BPXSY1)`
`nhanesSearch("glucose")`	`NHANES.search("glucose")`
`nhanesSearchVarName("BMXLEG")`	`NHANES.search_var_name("BMXLEG")`
`nhanesSearchTableNames("BMX")`	`NHANES.search_table_names("BMX")`
`nhanesDXA(2005)`	`NHANES.dxa(2005)`
`nhanesDXA(2005, suppl=TRUE)`	`NHANES.dxa(2005; suppl=true)`

Where the two differ

Both translate coded values by default, but they disagree on continuous variables. nhanesA converts a column to character as soon as the codebook reserves any code, so age, income ratio and the survey weights come back as strings. NHANES.jl leaves those columns numeric and skips translation for them, which means a weighted analysis works on the frame as returned.

The cost is that reserved codes stay in the column as numbers. See Value translation for how to find and filter them.

Component names accept the R short forms:

:DEMO or :Demographics
:DIET or :Dietary
:EXAM or :Examination
:LAB or :Laboratory
:Q or :Questionnaire