calc_group_stat {metacoder}

Apply a function to groups of columns

Description

For a given table in a taxmap object, apply a function to rows in groups of columns. The result of the function is used to create new columns. This is equivalent to splitting columns of a table by a factor and using apply on each group.

Usage

calc_group_stat(
  obj,
  data,
  func,
  groups = NULL,
  cols = NULL,
  other_cols = FALSE,
  out_names = NULL,
  dataset = NULL
)

Arguments

obj

A taxmap object

data

The name of a table in obj$data.

func

The function to apply. It should take a vector and return a single value. For example, max or mean could be used.

groups

Group multiple columns per treatment/group. This should be a vector of group IDs (e.g. character, integer) the same length as cols that defines which samples go in which group. When used, there will be one column in the output for each unique value in groups.

cols

The columns in data to use. By default, all numeric columns are used. Takes one of the following inputs:

TRUE/FALSE:

All/No columns will used.

Character vector:

The names of columns to use

Numeric vector:

The indexes of columns to use

Vector of TRUE/FALSE of length equal to the number of columns:

Use the columns corresponding to TRUE values.

other_cols

Preserve in the output non-target columns present in the input data. New columns will always be on the end. The "taxon_id" column will be preserved in the front. Takes one of the following inputs:

NULL:

No columns will be added back, not even the taxon id column.

TRUE/FALSE:

All/None of the non-target columns will be preserved.

Character vector:

The names of columns to preserve

Numeric vector:

The indexes of columns to preserve

Vector of TRUE/FALSE of length equal to the number of columns:

Preserve the columns corresponding to TRUE values.

out_names

The names of count columns in the output. Must be the same length and order as cols (or unique(groups), if groups is used).

dataset

DEPRECIATED. use "data" instead.

Value

A tibble

See Also

Other calculations: calc_diff_abund_deseq2(), calc_group_mean(), calc_group_median(), calc_group_rsd(), calc_n_samples(), calc_obs_props(), calc_prop_samples(), calc_taxon_abund(), compare_groups(), counts_to_presence(), rarefy_obs(), zero_low_counts()

Examples

## Not run: 
# Parse data for examples
x = parse_tax_data(hmp_otus, class_cols = "lineage", class_sep = ";",
                   class_key = c(tax_rank = "taxon_rank", tax_name = "taxon_name"),
                   class_regex = "^(.+)__(.+)$")

# Apply a function to every value without grouping 
calc_group_stat(x, "tax_data", function(v) v > 3)

# Calculate the means for each group
calc_group_stat(x, "tax_data", mean, groups = hmp_samples$sex)

# Calculate the variation for each group
calc_group_stat(x, "tax_data", sd, groups = hmp_samples$body_site)

# Different ways to use only some columns
calc_group_stat(x, "tax_data", function(v) v > 3,
                cols = c("700035949", "700097855", "700100489"))
calc_group_stat(x, "tax_data", function(v) v > 3,
                cols = 4:6)
calc_group_stat(x, "tax_data", function(v) v > 3,
                cols = startsWith(colnames(x$data$tax_data), "70001"))

# Including all other columns in ouput
calc_group_stat(x, "tax_data", mean, groups = hmp_samples$sex,
                other_cols = TRUE)

# Inlcuding specific columns in output
calc_group_stat(x, "tax_data", mean, groups = hmp_samples$sex,
                other_cols = 2)
calc_group_stat(x, "tax_data", mean, groups = hmp_samples$sex,
                other_cols = "otu_id")

# Rename output columns
calc_group_stat(x, "tax_data", mean, groups = hmp_samples$sex,
               out_names = c("Women", "Men"))


## End(Not run)