quickSource: Quickly Creating BigML Sources

Description

Quickly Creating BigML Sources

Usage

quickSource(data, name = deparse(substitute(data)), header = !is.null(names(data)), locale = "en-US", missing_tokens = c("NA"), quote = "\"", trim = TRUE, flatten = TRUE, ...)

Arguments

data

A matrix or data frame containing data to upload to bigml.

name

A string giving the name of the source.

header

A logical value indicating whether to use the first row of data as a header row.

locale

A string indicating the desired locale.

missing_tokens

A vector listing strings that should be treated as missing.

quote

A string giving the quote character to use.

trim

A logical value indicating whether to trim white space.

flatten

A logical value indicating whether to flatten the response into a data frame.

...

Arbitrary named arguments that are passed on to formEncodeURL in order to create form-encoded URL options.

Value

category: numeric
code: numeric
content_type: character
created: character
credits: numeric
description: character
fields: data.frame (or list if flatten=FALSE)
file_name: character
md5: character
name: character
number_of_datasets: numeric
number_of_models: numeric
number_of_predictions: numeric
private: logical
resource: character
size: numeric
source_parser: list
status: list
tags: AsIs
type: numeric
updated: character

Details

quickSource will take its "data" dataframe argument and attempt to create an equivalent BigML source. It does this by converting the dataframe to a csv file, compressing it, and uploading it directly to BigML. Generally, it's better to use quickDataset, since this method attempts to preserve any type information in the data frame.

References

https://bigml.com/developers/sources