Changed parallelization backend to use future and doFuture package instead of parallel package. This allows for more optimal parallelization on all operating systems. All parallelizations now use dynamic load balancing.
Removed Bigmemory version of find_exam function due to incompatibilities and stability issues.
Added progress bar to find_exam function to allow users to track progress of the function, even when parallelization is used.
Small bug fixes not affecting functionality.
Added export_notes function which provides a convenience function to write out each report as a separate text file.
Small bug fixes not affecting functionality.
Now support all data formats!
Simplified parallelization to only PSOCK
All functions now have unit test. Test coverage of code is 100%
create_img_db extensions now include: “.tmp”, “”
Suppressed start up messages in parallel loops
Added research_invitations variable to Con tables
Corrected Diagnosis_Flag to Procedure_Flag for Prc data
All columns are returned as is, no punctuation etc. is removed as previously
Changed load_all function to load_all_data, as now RPDR supports Allergy tables with the load_all function
In convert functions “time_type” arrgument is changed to “aggr_type” to support other aggregations methods in the future.
Removed all example datafiles.
nThreads for every function is now parallel::detectCores()-1 by default.
reticulate and bigmemory packages have been moved to Suggests. Importing of pydicom is now delay_load = TRUE. It allows users to specify a desired location for Python before interacting with your package.
load_notes now has an additional argument. If format_orig is set to TRUE, then the returned report maintains the original formatting, as opposed to FALSE, when all white spaces used for formatting are removed.
Bug-fix: file names and extensions are now converted to lower case for create_img_db to avoid mismatches due to capitals. Furthermore, the default extensions are now: “.dcm”, “.dicom” and “.ima” to avoid matches withing other parts of the file name string.
Bug-fix: in create_img_db now ““ is supported for identification of DICOM files without any extension.
Added load_report attribute to load_notes to specify whether the report text should be returned or not. Also, this attribute now affects whether load_all loads report text in case of notes.
Bug-fix: load_notes error where inappropriate partitioning of the file caused read errors.
Bug-fix: if data only contains 1 row, and identical column value removal is requested, the data.table is now returned.
Added create_img_db function which parses DICOM header information from radiological image series to create a database with meta-data.
Added lab_result_abn_flag_pretty column to convert_lab which marks abnormality based-on the Abnormal_Flag column in RPDR in lab.txt.
Bug-fix: problems with find_exam when data.table is small and when sequential execution is requested is fixed.
Now support Lno and Phy datasource.
Now support processing of Phy datasource using convert_phy function. Lno can be processed using the convert_notes function.
Updated README, and removed vignette.
Encounter numbers and accession numbers have been moved to the end of the data.table to make accessing them easier using indexes.
Inconsistencies in naming have been updated. Rdt and Lab tables time variable is simply: time_lab or time_rdt. Encounter data table’s encounter number is now enc_enc_numb. Time is now always in from in the naming, even for convert functions.
Bug fix: load_notes now names the new columns, based-on the type of note.
Bug fix: corrected load_prc in case there were carriage returns present in the name field.
Bug fix: added load_all compatibility with Dia datasource if date constraints are used.
Corrected documentation where needed.
Now support Prc datasource.
Now support processing of Prc datasource using convert_prc function.
Corrected documentation where needed.
Optimized sequential execution by not creating a socket when only 1 iteration is needed in convert functions.
Now support Rfv and all note type files: Car, Dis, End, Hnp, Opn, Pat, Prg, Pul, Rad and Vis using a single function: load_notes. Therefore, previous note related load functions have been removed.
Added convert_rfv functionality.
Now all notes files may be parsed using the convert_notes function. Therefore, previous note related convert functions have been removed.
For all datasources Clinic and Hospital names are not modified in any way so they are identical to the ones in RPDR, only missing values are changed to NA.
In functions parsing notes, exact matches are used for efficiency.
Bug-fix: load_dia did not load the encounter number of the record.
Bug-fix: parsing of notes has changed as in case of large files it caused memory issues.
Bug-fix: all columns are read as character as large MRNs interpreted as integers may cause problems.
Upon attaching the package the future.globals.maxSize is set to 100Gb to allow for parallel processing of large datasets on windows machines and also when using the shared memory version of the find_exam function.
Upon attaching the package the data.table setDTthreads is set to 1, so that there are no problems for parallel processing. Set data.table::setDTthreads() manually after the you are done with using parseRPRD, if you wish to use multi-thread processing of data.table in other tasks.
Timezone attributes are removed from POSIXct times as they may cause problems if working on different OS in different timezones.
In the find_exam function time variables are first truncated to the precision of time_unit argument using the trunc.POSIXt function for consistency, and then time differences are calculated using the difftime function.
Bug-fix: In case lab values were recorded as having “>” or “<” signs and the value was equal to the normal range i.e. value: <0.01, normal range: <0.01, then these values are considered as NORMAL.
Technical-fix: removed .. calls from functions so that only simple calls are present.
Need a high-speed mirror for your open-source project?
Contact our mirror admin team at info@clientvps.com.
This archive is provided as a free public service to the community.
Proudly supported by infrastructure from VPSPulse , RxServers , BuyNumber , UnitVPS , OffshoreName and secure payment technology by ArionPay.