dbosak01 / fmtr Goto Github PK

An R package for formatting data.

R 33.71% JavaScript 0.05% C++ 0.23% HTML 66.00%

r formatting fmtr format formats

fmtr's Introduction

Introduction to fmtr

R has a variety of ways to format data. There are functions for formatting dates, several functions to format numbers, plus you can always do a lookup with a named vector, or write your own vectorized function. Yet there are several problems with the way R handles formatting.

One problem is that there are too many formatting functions. It is not easy to remember which functions do what, and the differences between them. One goal of the fmtr package is to consolidate R's formatting capabilities into a smaller number of functions.

Another problem is when you want to reuse a format. While there are many ways to format data, there is no mechanism to save a format and reuse it later. Therefore, another goal of the fmtr package is to promote format reuse.

A third problem is when you have a complex format that can only be accomplished with multiple formatting functions. It would be nice to have a way to keep these multiple formats together, and assign them all at once.

Finally, there are some formatting activities that are so common, it seems like there should be a function for it. The fmtr package also contains several functions that fall into this category.

Solution

The fmtr package was developed to address the above problems. The package addresses those problems in the following ways:

Two formatting functions: With fmtr, all formatting has been consolidated into two formatting functions: fapply() and fdata(). fapply() applies a format to a vector. fdata() applies formats to a data frame or tibble. The implementation of the fapply() function is reminiscent of a SAS® put function.
The format catalog: The fmtr package also introduces the concept of a format catalog to R. A format catalog is a collection of formats that can be saved as a file, shared, and reused. This concept was taken directly from SAS® software.
User-defined format: The fmtr package also improves on the native R formatting capabilities by introducing the concept of a user-defined format. A user-defined format is an expression-driven lookup. You can define a set of logical expressions that, when true, will return a corresponding lookup value. The implementation of the user-defined format is similar to a SAS® user-defined format.
The formatting list: For complicated formats that require more than one function, the fmtr package also introduces the concept of a formatting list. A formatting list can perform operations that are difficult to accomplish otherwise.
Helper and Convenience functions: Finally, the fmtr package contains a set of functions to help you with common formatting tasks. For instance, it has functions to assist in assigning formatting attributes to an entire data frame. It also has some ready-made functions to perform very common types of statistical formatting.

All together, the above capabilities make formatting with the fmtr package both simpler and more powerful. The reuse features also make R more suitable to team programming.

For additional reading, examples, and a complete function reference, refer to the fmtr documentation site here.

Installation

To install the fmtr package, you may do so by running the following command from your R console:

install.packages("fmtr")

Then put the following line at the top of your script:

library(fmtr)

The fmtr package will then be loaded, and available for use in your project.

For examples and usage information, please visit the fmtr documentation site here. These examples will demonstrate the extraordinary usefulness of the formatting functions, and give you many ideas on how and where to use the fmtr package.

Getting Help

If you need help with the fmtr package, the best place to turn to is the fmtr web site. This web site offers many examples, and full documentation on every function.

If you want to look at the code for the fmtr package, visit the github page here.

If you encounter a bug or have a feature request, please submit your issue here

fmtr's People

Contributors

Stargazers

Watchers

Forkers

armenic

fmtr's Issues

Add SAS format names

Add translations of common SAS format names.

Error on as.fmt when input data is a tibble

as.fmt is failing when input data frame is a tibble. Reason is because bracket subsetting doesn't work the way it does with data frames.

User-define format errors when values not assigned

User defined format is bombing when a value "falls through" the format and is not assigned to a label. Enhance to allow values to fall through unaltered.

fapply() on user-defined format is failing when the conditions don't catch all vector values.

If a user-defined format doesn't catch all values in the vector, the format is failing. Would like any un-trapped values to fall though unaltered.

Allow assignment of NA to convenience functions

Allow user to assign an NA value to the convenience functions to clear attributes.

Limit rows of printed fcat output

Put some sort of default limit on the printed output of a format catalog. When it is large, it is returning 200+ rows, and is hard to read. Add a parameter to increase rows returned as desired.

as.data.frame.flist doesn't work right

as.data.frame.flist is generating a messed up data frame.

Separate format for SD

Add a parameter to the fmt_mean_sd function so that the use can format the standard deviation independently of the mean.

Add FAQ

Publish on Bioconductor

Publish this package on Bioconductor for greater visibility and accessibility.

Read SAS format catalog

Investigate reading a SAS format catalog and convert to a fmtr format catalog.

Check that fmt_cnt_pct is utilizing format code width

Allow fapply to accept non-character vectors

Sometime I want to map a character to a number or a logical. Allow fapply to do that.

Create pkgdown site

Create a pkgdown site for fmtr.

Allow user to format SD independantly

On fmt_mean_sd add a second format parameter to format the standard deviation separately from the mean.

Characters as factors on 3.6

On R 3.6, fdata() function is turning some character columns into factors. This is messing up the reporter package.

Vector format missing element

If a vector format doesn't have all values in data, missing items appear to be removed.

Widths issue

Not sure widths attributes are working properly on fdata. Don't seem to be applying to data frame.

Add description attribute

Add description attribute to fattr() function and create descriptions() function to assign all descriptions.

Numeric string format returning "NA" string

A numeric string format returns NA as "NA". This is very inconvenient, and messes up proper identification of NA values. Please make the NA fall through unaltered.

Import from SAS format file

A function to import a SAS format catalog from the format export file structure.

Improve documentation

as.fmt() function is not documented well. Put an example of using that, and creating a format from a data frame.

Clean up documentation

See some examples with "format()" function instead of "fdata()".

Review all documentation

Go over all documentation for accuracy and completeness.

Import function

Some sort of import function to make it easier to import codelists from a spreadsheet. Would have parameters to map the column names to what is necessary for the format. Returns either a data frame or a finished format catalog.

fdata() function is returning data type "As.Is" on some columns.

When you run the fdata() function on a data frame, it is returning the data type of most columns as "As.Is". This behavior seems undesirable. It would be preferable to leave the data type as was specified on the original data frame.

Possibly add label function

Existing functions to modify label attributes on a data frame are not sufficient. Consider adding a function to do that.

Add label parameter to fattr

Add the label attribute to the fattr function.

Add Complete Example

Add description attribute.

Add the description attribute to the set of formatting attributes. This needs to be done because there is no other package that supports a description attribute, and yet it is so commonly needed.

Support old R versions

Add GitHub actions to test with old R versions. Will need a few changes to codes, as this package uses some R 4.0 functions.

Vignette not working

Vignette not working for this package when calling the vignette() function.

Error on assigning fcat to df

When assigning an entire format catalog to a data frame using the formats() function, currently get an error if a variable is missing from the df that exists in the format catalog. Very inconvenient. Make it so that any variables that don't exist are ignored.