stats helpers {agsemisc} | R Documentation |
a.findcorrelations(df, vars1, vars2, min.cor
checks pairs of variables for correlation.
Returns a sorted vector of pairwise correlations whose absolute value
is greater-or-equal min.cor
.
Uses all pairs of variables from dataframe df
mentioned in
character vectors (vars1
x vars2
),
except that is returns only the lower
triangle of that matrix if vars2
is missing.
Assesses the variable itself and also the corresponding rank vector
(for rank correlations)
Value pairs containing NAs
are ignored in each cor
call individually.
a.iqr(x)
computes the inter-quartile range of numeric vector x
.
a.proportion.test(x1, x2, y1, y2)
tests the proportion x1/x2 against y1/y2 using both Fisher exact p and
Chi-squared test, prints results of both and returns results of Chi-squared.
If totals=TRUE
, tests x1/(x2-x1) against y1/(y2-y1) instead.
The arguments must be scalars, not vectors.
a.qr(x)
computes the quartile ratio (q75/q25) of numeric vector x
.
a.rankval(x)
ranks the values from vector x
using rank()
but leaves NAs
as NAs
.
a.showextremes(df, vars, largest, showalso)
shows the records from data frame df
that contain the
largest
largest values (or the largest
smallest if
largest
is negative) with respect to each of df
's variables
whose name is mentioned in vars
(which is a character vector).
The extreme records for each of these variables are found; from each
record, only these variables plus those whose name is mentioned in
character vector showalso
are shown.
a.findcorrelations(df, vars1=names(df), vars2=vars1, min.cor=0.5) a.iqr(x) a.proportion.test(x1, x2, y1, y2, totals=FALSE) a.qr(x) a.rankval(x) a.showextremes(df, vars, largest=5, showalso=NULL)
df |
a dataframe, see description above. |
vars1 |
a character vector, see description above. |
vars2 |
a character vector, see description above. |
min.cor |
a number in range 0..1, see description above. |
x |
a vector, see description above. |
x1 |
see description above. |
x2 |
see description above. |
y1 |
see description above. |
y2 |
see description above. |
totals |
a logical value, see description above. |
vars |
a character vector, see description above. |
largest |
an integer, see description above. |
showalso |
a character vector, see description above. |
Lutz Prechelt prechelt@inf.fu-berlin.de
data(iris) summary(iris$Sepal.Width) a.iqr(iris$Sepal.Width) a.qr(iris$Sepal.Width) a.findcorrelations(iris) a.proportion.test(7,11, 17,19) a.showextremes(iris, c("Sepal.Width","Sepal.Length"), largest = -5, showalso = c("Petal.Length"))