Changes in version 1.2.21 - Add validation check for spectrum_region_code in the shape file. It must: - Be set for every feature - Be all 0 or if for a country that uses subnational PJNZ have NA for national level and be unique for every level 1 area. Changes in version 1.2.20 - Improve error message when a worker crashes. Changes in version 1.2.19 - Fix issue where worker config extra memory option was being parsed incorrectly Changes in version 1.2.18 - Update to save files in qs2 format instead of qs. qs package has been deprecated. - Update metadata for indicators and filters to show time series data in input choropleth. Changes in version 1.2.17 - Improve docs for download_debug, change token name and update order of args Changes in version 1.2.16 - Update Naomi model options for 2026 estimates Changes in version 1.2.15 - Update download_debug to use GitHub token for auth and pull from new authenticated endpoint. Changes in version 1.2.14 - Remove endpoints /rehydrate/submit, /rehydrate/status/ and /rehydrate/result/ in favour of moving these to Kotlin. - Add endpoint /task//exists to check if a task with that ID exists. Changes in version 1.2.13 - Update worker configuration setup to accept a json configuration file instead of being hardcoded in the package. Changes in version 1.2.12 - Update ART comparison table data and metadata to return adjusted and unadjusted spectrum values. Changes in version 1.2.11 - Add new endpoint for downloading datapack output nm-103. Changes in version 1.2.10 - Remove ANC strict validation for Zimbabwe to be able to upload data Changes in version 1.2.9 - Return metadata from a new separate population endpoint to make re-fetching when languages change easier. See nm-94. Changes in version 1.2.8 - Return data and metadata from population validation endpoint Changes in version 1.2.7 - Return a more human readable error message when reading a csv fails in spectacular fashion. Changes in version 1.2.6 - Update test data and options to new Malawi 2024 data Changes in version 1.2.5 - Return metadata for input comparison table. Changes in version 1.2.4 - Add a new endpoint /chart-data/input-comparison for returning data and metadata for input comparison barchart. Changes in version 1.2.3 - Add checks that a rehydrated project is valid. It checks that: - All input files exist on disk - The model fit and calibrate IDs are known in redis Changes in version 1.2.2 - Fix issue where if shape file was uploaded with no display property it was raising an error when fetching input metadata. Now if no display property is set, it will include it in the returned metadata. Changes in version 1.2.1 - Make it harder to accidentally delete redis data by: - Not starting workers automatically when you create a queue - Only removing the workers a queue creates by default - Only deleting data if you explicitly opt in - Add new endpoint /download/result/path/ to return path to the download file on disk Changes in version 1.2.0 - Update plot metadata format to make it more generic and generally usable - Add a new endpoint /review-input/metadata to return metadata for all input plots - Remove /meta/plotting and /meta/plotting endpoints as these are no longer used Changes in version 1.1.25 - Use new rrq functions Changes in version 1.1.24 - Return calibrate result path from /calibrate/result/path/ relative to the results directory Changes in version 1.1.23 - Update to use new rrq free functions added in v0.7.10 Changes in version 1.1.22 - Remove hardcoded year2 value for ANC prevalance and ANC ART coverage Changes in version 1.1.21 - Update tableMetadata to return id and label for row and column values. Changes in version 1.1.20 - Add tableMetadata into the calibrate/result/ and calibrate/result/metadata/ endpoints for controlling how the new output table feature will look Changes in version 1.1.19 - Ensure that plot data in submit_prerun is uploaded as a duckdb file. Changes in version 1.1.18 - Add a function to migrate from naomi v 2.9.10 to 2.9.11. i.e. to migrate from plot data being stored as .qs files to being stored as duckdb database. Changes in version 1.1.17 - Add a dummy download endpoint of type "agyw" for generating AGYW (adolescent girls and young women) tool Changes in version 1.1.16 - Add an optional arg to API startup to turn on a health-check. If set, the interval after which the next time the redis connection is used the connection will be reset. 0 for no reconnection. This is needed for running on cloud services as they close idle tcp connections after a set time. On Azure it if 4 mins, so we need to the client to be able to reconnect if this interval has passed. - Add a new endpoint /calibrate/result/path/ which will return the path to the calibration plot data output which can be read by web backend Changes in version 1.1.15 - Add two new endpoints in preparation for web backend accessing the result data on disk directly - /calibrate/result/metadata/ to return the metadata and warnings from the /calibrate/result/ endpoint - /calibrate/result/data/ for returning only the data from the /calibrate/result/ endpoint Changes in version 1.1.13 - Error early if reading data generates a warning, we cannot proceed with only a partial read of the data Changes in version 1.1.10 - Update fallback anc year to 2022 - Remove stack trace from user display - Return job_id in error response if error originated from async job Changes in version 1.1.9 - Ensure Naomi options are translated based on accept-language header Changes in version 1.1.8 - Add hintr_prerun function which takes model inputs and outputs, saves them onto the server and then returns output zip with state JSON which can be used to rehydrate model outputs Changes in version 1.1.7 - Remove prerun_import and prerun_push functions, these are going to be replaced with a more transparent prerun method Changes in version 1.1.6 - Add endpoint /internal/prerun to run a prerun which builds state JSON from input and output files Changes in version 1.1.2 - Add default selections for each indicator in comparison plot Changes in version 1.1.1 - Add fallback model and calibration options for unknown countries Changes in version 1.0.44 - Fail early if comparison report output cannot be generated Changes in version 1.0.41 - Add country specific default values to all model and calibration options Changes in version 1.0.39 - Add dummy /comparison/plot/ endpoint to return data for comparison barchart Changes in version 1.0.38 - Save project_state.json into info dir in the output zip to make it less visible to users Changes in version 1.0.37 - Add endpoints /rehydrate/submit /rehydrate/status/ and /rehydrate/result/ for getting state JSON required by the front end for rehydrating a project Changes in version 1.0.36 - Delete exited workers when API starts (so metrics on worker counts are up to date see mrc-2893) Changes in version 1.0.35 - Download submit for spectrum can take optional state JSON which is saved out as is into the output zip which can be used to by the web end to recover the project Changes in version 1.0.34 - Include input files when download_debug called with id from a calibrate run or output generation Changes in version 1.0.33 - /download/submit// can now take a body with notes which will be added to spectrum download zip, it is ignored for other download types. Changes in version 1.0.32 - Temporarily stop warning generation from ANC & ART uploads as it is taking a long time (> 18s for Nigeria) which is causing time out issues in hint. Changes in version 1.0.31 - Add new download type "comparison" to return input to output comparison report Changes in version 1.0.29 - Return warnings from input time series aggregation Changes in version 1.0.24 - Validate that spectrum file and geojson contains same spectrum region codes Changes in version 1.0.23 - Ignore non-pjnz files from uploaded zip file instead of erroring (so that mac users whose system creates __MACOSX file in zip files do not get errors) Changes in version 1.0.22 - Validate columns present before other validation checks Changes in version 1.0.18 - Return ratios as indicator types in calibrate plot Changes in version 1.0.17 - Add query param strict to endpoint /validate/survey-and-programme to allow running with relaxed validation, by default this will run in strict mode. - Add /meta/plotting endpoint to return default plotting metadata. Changes in version 1.0.15 - Update programme & ANC validation to ensure only 1 area level per year is uploaded Changes in version 1.0.13 - Add validation error to programme/ART data to ensure only 1 area level is uploaded Changes in version 1.0.12 - Return warnings from model option validation, model fit and model calibrate endpoints Changes in version 1.0.11 - Return format and accuracy metadata for data_type columns in input time series to specify how data of that type should be shown in the front end. Changes in version 1.0.10 - Recode vls_tested and vls_suppressed to vl_tested_12mos and vl_suppressed_12mos. Affects tests only. Changes in version 1.0.9 - Back /calibrate/plot/ endpoint with real data Changes in version 1.0.8 - Remove time_step from input time series endpoints and add quarter Changes in version 1.0.6 - Return area and time_period column info in input time series metadata Changes in version 1.0.1 - Remove PJNZ file from survey and programme data validation - Wire up /chart-data/input-time-series/ to return real data for time series plots Changes in version 1.0.0 - Update downloads to be generated on demand instead of as part of model calibration Changes in version 0.1.39 - Add dummy endpoint /chart-data/input-time-series/ to return dummy data for upcoming input time-series plots Changes in version 0.1.38 - Add Portuguese translations Changes in version 0.1.37 - Add dummy endpoint /calibrate/plot/ to return dummy data for upcoming calibrate plot Changes in version 0.1.36 - Update to latest version of rrq Changes in version 0.1.35 - Shorted output file names Changes in version 0.1.34 - Remove deprecated /model/calibrate/ endpoint Changes in version 0.1.33 - Return upload metadata from calibrate result response Changes in version 0.1.32 - Update summary report Changes in version 0.1.31 - Run model and calibration in a separate process to avoid memory leak Changes in version 0.1.30 - Update tests for naomi 2.3.0: add new model option calibrate_method. Changes in version 0.1.29 - Update to naomi 2.2.4, save calibrate outputs to new file Changes in version 0.1.28 - Turn off validation for survey data n_cluster column, this is no longer required Changes in version 0.1.27 - Add arg to worker entrypoint to enable starting calibration only worker Changes in version 0.1.26 - Update test data options to new demo survey ids "DEMO2015DHS" and "DEMO2016PHIA". Changes in version 0.1.25 - Queue calibration with higher priority than model run Changes in version 0.1.24 - Add endpoints to enable async calibration - /calibrate/submit - queue a model calibration - /calibrate/status - get status of a calibration run - /calibrate/result - get result of a calibration - Add calibration options back in and remove from run options Changes in version 0.1.23 - Update to naomi v2.1.11 (branch pin naomi@naomisup-ticket-11). Changes in version 0.1.22 - Update to naomi v2.1.10 (branch pin naomi@infections-metadata). Changes in version 0.1.20 - Pin to naomi@eppasm-0.5.9 Changes in version 0.1.18 - Update summary report Changes in version 0.1.17 - Pin to naomi@aware-of-status Changes in version 0.1.15 - Pin to naomi@option-output-aware Changes in version 0.1.14 - Generalize functions assert_calendar_quarter_column() and assert_year_column() to accept argument col_name= to check columns of required specification with different name. Changes in version 0.1.13 - Pin to naomi@issue-142 Changes in version 0.1.12 - Pin to naomi@anc-testing-cascade Changes in version 0.1.11 - Add endpoint <+calendar_quarter_t1_default+> to select most recent survey calendar quarter as default 'time 1' option. Changes in version 0.1.9 - Return helpText from model run Changes in version 0.1.8 - Revise ART programme data schema to use column calendar_quarter instead of year. Changes in version 0.1.7 - Depend on naomi 2.0.0 version for 2021 UNAIDS estimates. Changes in version 0.1.6 - Reflect pkgapi rename Changes in version 0.1.5 - Add endpoint for downloading dummy summary report Changes in version 0.1.4 - Return number formatting metadata from /meta/plotting/ and /model/result endpoints Changes in version 0.1.3 - Add model calibration endpoint Changes in version 0.1.2 - Rename /download/summary endpoint to /download/coarse-output to be more representative of actual content and in preparation for adding an endpoint for downloading summary report Changes in version 0.1.1 - Fix file naming to work with old model runs where country name is not available Changes in version 0.1.0 - Switch API implementation to use pkgapi Changes in version 0.0.36 - Return uncertainty range metadata from model result for choropleth Changes in version 0.0.34 - Accept semicolon delimited data as csv inputs (common in countries with non-Anglo number formats) Changes in version 0.0.33 - Language preference is passed through to naomi during model runs. Changes in version 0.0.32 - New debug endpoint for locally getting a copy of failing data. Changes in version 0.0.31 - Extend time options back to Q1 2010 to support household survey datasets back to 2010 (Burkina Faso). Changes in version 0.0.30 - Remove "survey_year" from required columns in survey datasets. Changes in version 0.0.28 - Add default values for area level and calendar quarter to generate estimates for Changes in version 0.0.27 - Support for stopping jobs (mrc-732) Changes in version 0.0.25 - Add receiving_art output indicator Changes in version 0.0.24 - Return better error message from geojson reading Changes in version 0.0.23 - No longer require ancrt_hiv_status - Internationalisation support (mrc-788) Changes in version 0.0.22 - Update for changes to model options in naomi v0.0.21 - Support prerun model results (mrc-1155) Changes in version 0.0.21 - Errors in the hint model run are returned with stack traces (mrc-714) Changes in version 0.0.20 - Accept and validate zip of PJNZ files Changes in version 0.0.19 - Return version info from model run options response - Require version info in model run submit endpoint - Update filter ordering and naming Changes in version 0.0.18 - Add model options validate endpoint - Update input data structures including - Move quarter_id to use calendar_quarter instead in format CY2016Q3 for population and output data - Move quarter_id to use year for ANC, ART, - Move age_group_id to age_group format e.g. 00-04, 15+, 05-09 etc. in ANC, ART, population, programme, survey and output data - Update model run options - Update region IDS to be formatted as e.g. MWI_2_1 for first region in level 2. Top level region remains as just ISO3 code e.g. MWI - Update filters and metadata to reflect data changes - Add advanced model run options Changes in version 0.0.17 - Not found (404) responses now conform to the error schema (mrc-596) Changes in version 0.0.16 - Download endpoints return Content-Disposition headers Changes in version 0.0.15 - Add HEAD endpoints for summary and spectrum downloads Changes in version 0.0.14 - Return plotting metadata for barchart with the model result response Changes in version 0.0.13 - Caching enabled for geojson reading Changes in version 0.0.12 - Update model download endpoint from indicators to summary - Get downloads from model run - Replace model run code with a mock that returns expected format - Allow running up hintr docker container with env var USE_MOCK_MODEL to mock a model response Changes in version 0.0.11 - Return iso3 code from PJNZ upload - Use iso3 to locate plotting metadata instead of country name Changes in version 0.0.10 - Iterate model run option endpoint to return complete options, to always return an ID and label for each option and to return regions as a hierarchy Changes in version 0.0.9 - Add stub endpoints for downloading key indicators and spectrum digest Changes in version 0.0.8 - Add endpoint for retrieving model run options Changes in version 0.0.7 - New GET /hintr/version and GET /hintr/worker/status endpoints, the first in a series of hintr informational endpoints Changes in version 0.0.6 - Fix plotting metadata for output dataset - Return indicator filters for input data - Make input and output data responses more consistent Changes in version 0.0.5 - Add endpoint for retrieving plotting metadata Changes in version 0.0.4 - Rename validate/input endpoint to validate/baseline-individual and rename validate/baseline to validate/baseline-combined for clarity Changes in version 0.0.3 - Validate survey and programme data using a different endpoint from the validation of baseline data Changes in version 0.0.2 - Add baseline validation endpoint and update validation endpoint names Changes in version 0.0.1 - Initialise package