rayamberg / autopatt Goto Github PK

View Code? Open in Web Editor NEW

3.0 1.0 1.0 109 KB

Phon script to automate the Phonological Assessment and Treatment Target (PATT) selection process.

License: Other

Groovy 100.00%

autopatt's Introduction

AutoPATT

Phon script to automate the Phonological Assessment and Treatment Target (PATT) selection process.

INSTALLATION INSTRUCTIONS:

For the latest script (requires Phon 2.2 or above):

Click Code on upper-right of project toolbar.
Click Download .ZIP file
Extract zip file
Move AutoPATT.groovy to your Phon scripts folder. By default this will be in ..\Documents\Phon\PhonShell or ... ~User\AppData\Roaming\Phon\PhonShell, depending on your version. If this is your first script, you will need to create the "PhonShell" folder.
AutoPATT should now appear in Phon when you click Plugins->PhonShell scripts

For older versions of Phon:

Please navigate to Releases and download the .ZIP file of AutoPATT that is compatible with your version of Phon. This will be described in the release notes for AutoPATT.
Extract zip file
Move AutoPATT.groovy to your Phon scripts folder. By default this will be in ..\Documents\Phon\PhonShell.
AutoPATT should now appear in Phon when you click Plugins->PhonShell scripts

autopatt's People

Contributors

Stargazers

Watchers

Forkers

philcombiths

autopatt's Issues

Handle mismatches between orthography and IPA Actual

User should be given an option when there does not appear to be a good match between the orthography, which helps to map meaning to minimal pairs, and the actual IPA transcription. Perhaps a warning or option to skip.

Add disclaimer about AutoPATT limitations

AutoPATT does not consider stimulability for English/Spanish PATT and does not consider s+nasal clusters in English.

In the CSV output, include speaker, session, and language information.

Ensure affricates are not being considered in the sonority distance calculations.

Misidentification of dipththong as fricative in phonemic inventory

a͡ʊ identified as a fricative in phonemic inventory from Damon dataset.

ma͡ʊ | | mo͡ʊ

Check to ensure that minimal pair rule 5 is followed

Do not repeat minimal pairs. If you use [maUs] and [maUT] for /s/, don’t use the same minimal pair
for /T/, unless there are no other minimal pairs found in the data for /T/.

May or may not decide to implement this depending on how "strict" we want to be with minimal pairs

Visitor Pattern for CSV writing

Consider implementing a visitor pattern that visits each key component and writes that way. This visitor will need information about which language is being processed and its rules. For example, when visiting an English phonetic inventory, it should know the basePhones for English, and when visiting Spanish (or Russian, etc.) it will have to print things out differently.

Consider contrasting allophones

Allow for more narrow transcription.

Present CC cluster targets: First with 2 sounds missing, then with 1 sound missing. Do we consider clusters with no sounds missing?

Present CC cluster targets: First with 2 sounds missing, then with 1 sound missing.

Do we consider clusters with no sounds missing?

Step 2 is indicating possible clusters with only one out phone, but it should be that both phones in the cluster are out.

The issue should be resolvable at line 473. Make both elements in the cluster pass the if statement requiring both to be out phones

add instructions for stimulability for the user

Difference between regular "g" and IPA "ɡ"

If a user types in a "g" this will lead to unexpected behavior. Some kind of check for the IPA /ɡ/ must exist within the script.

CSVWriter does not close open file upon completing script

This bug does not occur in older versions of Phon (e.g., v2.2) but does occur in Phon v3.1

Add final clusters inventory

change encoding of output CSV

change encoding of output CSV so IPA symbols appear correctly in Excel. Current encoding is UTF-8. UTF-8 BOM would work. Or perhaps create an option to choose.

Corrections to CSV output

Modify arrangement of inventory tables in CSV output:

[h] as glide in English,
Change row order (fricatives before nasals),
Change row labels: liquids instead of laterals, remove other fricatives

Spanish Inventory should have ɲ

Spanish inventory is currently monitoring the retroflex nasal, and needs to be monitoring ɲ

The minimal pair meaning contrasts don't account for homonyms

The meanings map only stores one meaning per production. It should be storing a list of meanings since a child may produce the same sound for multiple meanings. This will create improper results if homonyms are used in the probe.

Re-design constructors for inventory classes

The buildInventory() and gather() functions may not be necessary if the proper functions are being executed in the constructors for each speaker. Consider doing all the parsing in the constructor.

.csv suffix not automatically appended to filename

User must type ".csv" to create a csv file. This should be done automatically.

Add phonemic inventory for Spanish PATT

Instability on Mac

There are a number of issues with AutoPATT on the Mac version of Phon v2.2. This problem was replicated on a Macbook running MacOS Sierra 10.12.5. Issues include the following:

Phon often crashes (not all the time) when executing the script initially
Script is unable to execute a second time after successful completion the first time. Phon crashes.

Force quitting and restarting Phon is usually enough to remedy the first problem, but the second problem persists.

Possibility to split PATT.groovy into multiple files

Right now all of the files are within one PATT.groovy script. It would be nice to be able to import the classes from separate files, but I can't seem to figure out how to do that within PhonShell/Groovy. Would this be possible?

Confirm that AutoPATT reflects PATT 2.0

When cluster targets are not found, display other less ideal clusters before singletons

Create menu of functions

Allow for menu of functions, so that user doesn't just have to do PATT, but could just get a minimal pair count, for example.

Include PATT references into script or output

The references and links to SLPath.com and others need to be included somewhere either in the script itself, or in the file output. Perhaps a separate PDF

PATT English Step 2, Part D should try to guess pattern

Current recommendation: It can spit out two lists, one of 2-element non-s clusters and another with 2-element s-clusters. The s-cluster list can come with a caveat, e.g. These may be appropriate targets if they are NOT error-patterning like /st-/, /sp-/, /sk-/ (violate SSP, are adjunct clusters). Also, this may be difficult, but if it could spit out a summary table of errors for the s/STOP/ clusters and the other s-clusters in questions. If they show the same reduction pattern, then do it. If one reduces to C1 and the other doesn't then they aren't patterning the same. (P1/P2)