NEWS.md

In tjebo/eye: Analysis of Eye Data

1.2.1

Bug fixes: - if no eye column is found, eyes also returns a list #31 - set_eye_strings also updates eye column #32 - partial eye strings recognized in column names #33 - to_etdrs now correctly always returns integer class vector #34 - conversion of pure "NA" saved as character now correctly returns NA #35 - logMAR will return rounded values to the first digit when noplus = TRUE #37 - to_etdrs from pure qualitative values correctly returns 0 for NPL and PL #38

1.2.0

Summary of changes: - eyes() - now returns lists (of class "eyes") for easier access of count data - set_eye_strings() - set string codes globally! This makes it easier for people using different languages to use eye. - recodeye: renamed "eyecodes" argument to "eyestrings" - Change of terminology to "subjects/id" rather than "patients" - simplified code

Details: - eyes(): "details" argument creates object of class "eyes_details", allowing access to in depth count statistics and subject ids - This includes new print methods for class "eyes" and "eyes_details" - change id and eye arguments to "id_col" and "eye_col" - previous code should not break because of partial argument name matching. However, the new names make it clearer as to what those arguments are for. - calling set_eye_strings will super-assign the new codes to the internal list object eye_codes - getElem_eyecol now gives precedence to columns that are called "eye" or "eyes" (#30) Simplifications: - simplified pivoting in myop using the pivot_longer regex feature - removed ...chr arguments from getElem... functions (redundant) - getElem... now only for vectors (because they were anyways only used for vectors)

1.1.0

summary of changes: - new features for function eyestr - important bug fixes

Details: - eyestr: new "english" argument for more flexibility (#26) - eyestr removed "para" argument - eyestr prints one eye correctly in singular (#25) - eyestr added "caps" argument for more flexibility of spelling - eyestr removed "UK" argument (there is no difference in spelling for numbers :) - eyes fixed missing reference to column (#27) - added internal function tidyNA_low to avoid double tolower call - recodeye fixed fail when trailing white space (#28) - which_va fixed conversion error when mixed ETDRS and logmar (#29)

eye 1.0.1

• Nasty bug fix! Erroneous conversion of quality visual acuity entries when ETDRS (#24)
• removed deprecated function age()
• VA chart now as internal data, not exported - ideally the VA chart should not be necessary anyways. However, if you really need to see it, you can still access it with eye:::va_chart. Note some of the ETDRS values were chosen for easier conversion.
• internally used S3 methods not exported to namespace any longer:
• convertVA
• checkVA

eye 1.0.0

Major version upgrade of eye 0.1.0!

Functions and objects (details)

Visual acuity handling

• introduced new verbs as_... and to_... for class conversion for nice integration into known R grammar (#6) - simple wrapper for va(x, to = ...)
• add arguments "from", "noplus" and "smallstep"
• removed "from_logmar" and "logmarstep"
• VA notation guessing is not done by single element any more, but will be forced choice either by most likely or with argument "from". Any implausible values will be forced to NA.
• weird snellen values (e.g., 20/41) are now also converted to other snellen values (#21).
• New S3 methods introduced to check plausibility of entries by VA class.
• Adding plus/minus entries is now done by actual logmar values rather than via lookup in the VA chart (rewrite of snellen_steps function)
• new function va_mixed for vectors of mixed VA notations (replacing previous va_dissect)
• new function cleanVA

Smaller fixes

• recodeye() recognises "both/ou/b" for both eyes.
• dropunknown unclear codings are converted to NA by default (with warning)
• removed numcode argument (other numeric codes can be passed via "to")
• eyes() accepting both quoted and unquoted arguments (#16)
• add stop() when main argument not a data frame
• eyes() now also deals with "both eye cases" (#17)
• now also finding eye columns with funny names, e.g. EyeName (#18)
• eyes now returns meaningful result on empty data frames) (#19)
• removed "report" argument
• reveal() does not drop unused group levels (#13)
• eyestr() moved eyes_to_str step from eyes() to eyestr()
• age() renamed to getage() in order to avoid potential and even likely clashes with the users environment, especially when using age() within the local environment of a data frame (e.g., using with())
• blink() is deprecated. message that this function will no longer be maintained, but kept in the package.
• isNAstring now also converts "-" string to NA
• new exported function: tidyNA() cleans NA entries
• clean_va returns numeric vector if all entries are numeric
• removed functions va_dissect() and which_va_dissect()

Data sets

• Removed data set amd - this will be now part of the dedicated ophthalmic data set package "eyedata"
• removed data set 'va_quali' which was anyways a part of "va_chart"

eye 0.1.0

the eye package is online! eye is a tool to facilitate common tasks in ophthalmic research.

Its visual acuity conversion function va() provides eye researchRs with a long coveted tool to handle different visual acuity notations. It works with Snellen, logMAR and ETDRS.

recodeye allows an easy recoding of the eye variable.

It also has some other functions which are mainly intelligent wrapper around already existing functionalities, such as myop and hyperop for reshaping eye specific data, reveal or blink.

See your data with a new eye.

tjebo/eye documentation built on Jan. 24, 2022, 8:34 p.m.