# itremove: Iteratively removes all rows and columns of a matrix or... In MartinHinz/quantaar: R Library for Quantitative Analysis in Archaeology

## Description

Iteratively removes all rows and columns of a matrix or dataframe with less than a given number of non zero elements

## Usage

 `1` ```itremove(x, minnumber = NA, cmin = NA, rmin = NA) ```

## Arguments

 `x` matrix or data.frame. Table with only numeric values. `minnumber` integer. A number of minimum non zero elements. How many values > 0 have to be present to consider a row/column sufficiently linked to perform further analysis? Every column with less values > 0 will be removed. Either `minnumber` or `cmin` and `rmin` must be given. `cmin` integer. Same as `minnumber`, but specifically for columns. `rmin` integer. Same as `minnumber`, but specifically for rows.

## Details

A matrix or data.frame with numeric values often contains rows and columns with an insufficient amount of values > 0 for a certain task. For example correspondence analysis or bivariate correlation analysis requires a minimum amount of usable values. In an archaeological context this could apply for example, if certain find categories are particularly rare in a burial site context.

`delrc` allows to remove rows and columns, that don't fulfill the requirements.

## Value

A matrix or dataframe with all rows and columns removed that had less than the given number of non zero elements. If `minnumber` or `cmin` and `rmin` are too restrictive and no content remains in the table, than the result is `NA`.

## Examples

 ``` 1 2 3 4 5 6 7 8 9 10 11 12 13 14``` ```testmatrix <- data.frame( c1 = c(0,3,8,2), c2 = c(0,6,7,0), c3 = c(0,0,0,0), c4 = c(0,3,8,2), c5 = c(0,6,7,0), c6 = c(1,0,0,1) ) # The following code removes every column with less than 3 values > 0. # That will remove the columns c2, c3, c5 and c6. # Further, every row with less than 2 values gets removed. # That will delete row 1. itremove(testmatrix, cmin = 3, rmin = 2) ```

MartinHinz/quantaar documentation built on May 7, 2019, 3:37 p.m.