\(\newcommand{\B}[1]{ {\bf #1} }\) \(\newcommand{\R}[1]{ {\rm #1} }\) \(\newcommand{\W}[1]{ \; #1 \; }\)
2024¶
View page sourceRelease Notes for 2024¶
mm-dd¶
12-02¶
Advance Ubuntu in the dismod_at.base image from 23.04 to 24.04 in the dismod_at.base image. This also advances python3.11 to python3.12 .
Advance dock_dismod_at to dismod_at_version to 2024.12.2 and at_cascade_version to 2024.12.2 .
The dock_dismod_at.sh Images documentation has been separated from Working with Images .
11-26¶
It is an error if there is a zero or negative weight value in the weight_grid table, and the corresponding weight_id also appears in the data_table or the avgint_table . The error message for this case has been improved. To be specific, the corresponding row of the data table or avgint table is reported.
10-20¶
Advance from cppad_mixed-20240910 to cppad_mixed-20241029.
10-16¶
Garland Culbreth found an error in the Bilinear formula. To be specific the second \(v_{i+1,j}\) was fixed by changing it to \(v_{i,j+1}\) . In addition, a table was added below that formula that checks the value at the vertex points.
09-24¶
It was an error for the prior table to have an eta that was zero and a density that was Log Scaled . This has been changes so that eta can be zero if lower plus eta is greater than zero.
Advance from cppad_mixed-20240902 to cppad_mixed-20240910.
Advance dock_dismod_at to dismod_at_version to 2024.9.24 and at_cascade_version to 2024.9.24 .
09-09¶
The development and install tools were extended to work well on
MacOS with homebrew.
The versions between 08-25 and 09-08
were not tested and may have trouble on other systems.
(There was a problem with the new bin/grep_and_sed.sh script.)
08-18¶
The specifications for errors in the log table now include the fact that the program exits with a non-zero exit code directly after reporting one of these errors.
08-13¶
Add read permission for the world to the
dismodat.pyfile so that now user, group and world can execute this script. (It used to be the case that only users in the same group as the installer could execute this script.)Advance dock_dismod_at dismod_at_version to 2024.8.13.
07-30¶
Move the install settings from
bin/run_cmake.shto bin/install_settings.py . Theinstall_settings.pyroutine is better able to transfer results that depend on the system it is run on.Remove the
bin/ihme_db.shscript (it was out of date, was IHME specific, and would no longer be useful).Advance dock_dismod_at dismod_at_version and at_cascade_version to 2024.7.30.
07-29¶
Advance from cppad_mixed-20240214 to cppad_mixed-20240729.
07-15¶
Fix a compiler warning in get_option_table_xam.cpp. (This is a test and developer example.)
07-10¶
Advance dock_dismod_at dismod_at_version and at_cascade_version to 2024.7.10. This tests the changes on 07-09 below (and fixed a problem introduced in dock_dismod_at.sh).
07-09¶
Change the install of the dismod_at python module so that it automatically includes any packages it depends on. This is should solve some problems when using example_install.sh with homebrew on macOS.
06-28¶
There was a bug in the most recent change to dock_dismod_at.sh.
(It would get an error about changing bin/run_cmake.sh when attempting to
checkout out a specific version of dismod_at .)
This has been fixed.
06-27¶
Fix some problems with the install of dismod_at on macOS.
To be more specific, example_install.sh should now work
if you use homebrew as a package manager and install the
following packages numpy, python-matplotlib , scipy
before running example_install.sh.
06-24¶
Advance dock_dismod_at dismod_at_version to 2024.6.23 and at_cascade_version to 2024.6.24.
06-22¶
Add subgroup random effects to the covariate multipliers in the user_zero_meas_value.py example. In addition, fix the setting of these random effects to zero when the zero_meas_value option is specified.
06-21¶
Add the zero_meas_value option to the predict command. In addition, add the user_zero_meas_value.py example.
06-10¶
Fix usage message for dismodat.py . To be specific, change
perturb to perturb .
05-25¶
Add the Version command.
Advance dock_dismod_at.sh to dismod_at-2024.5.25 and at_cascade-2024.5.25.
05-22¶
Change the dismod_at version numbers from yyyymmdd format to year.month.day format; e.g, the old version number for 2024-05-22 is 20240522 and it’s new version number is 2024.5.22 .
Advance dock_dismod_at dismod_at_version to 2024.5.22 and at_cascade_version to 2024.5.20. In addition, fix using same dismod_at mixed image when only dismod_at or at_cascade changes.
05-21¶
Fix a misspelling of driver (as driver) in dock_dismod_at.sh.
03-12¶
The dock_dismod_at.sh had the version of dismod_at baked in. This has been changed so that it is no longer necessary to re-build the base image each time the version of dismod_at changes.
Advance dock_dismod_at.sh` to dismod_at-20240310 and at_cascade-2024.3.12.
03-10¶
The create_connection would sometimes hang when used with python multiprocessing. The connections have been changed so that they can only be used by One Thread . This seems to have fixed this problem.
02-20¶
The build_type was fixed in the base image, instead of the mixed image. This is has been fixed, in addition, all the images double check that they are building the correct version and build type.
Change the page name
whats_new_year to just year . This shortens the links to the dismod_at release notes; e.g. , the link:https://dismod-at.readthedocs.io/whats_new_2024.html#whats-new-2024-mm-dd-02-14
changed to:
https://dismod-at.readthedocs.io/2024.html#mm-dd-02-14
02-18¶
The user_diabetes.py example was completely re-written. This has been improved in many ways. For example, it is simpler than other examples that simulate data because it uses average_integrand , instead of the simulate_command .
If one uses a rate_eff_cov_table , and they have a lot of nodes and covariates, setting up the interpolation for this covariates can be time consuming. This is now only done for the commands that need this information. This changed the execution time for the init_command or hold_out_command commands from about 30 minutes to about 12 seconds (for one IHME Mexico diabetes estimation case).
Advance dock_dismod_at dismod_at_version to 20240219 (to include the changes above).
02-17¶
There were two rows with the value of ode_step_size in the db2csv option.csv file. This has been fixed.
02-14¶
Make n_data an adjustable parameter to the speed.py timing tests. Include the hold_out_command in the speed.py timing tests. In addition, separate the elapsed time into the different steps taken by speed.py.
Advance to cppad_mixed-20240214. This fixes a bug in CppAD, which was discovered using the new n_data argument and hold_out commands in speed.py .
01-31¶
Add a hold out example to the Batch Command wish list item.
01-28¶
In dock_dismod_at.sh advance to dismod_at-20240105 and at_cascade-2024.1.28.
01-05¶
API Change : The dock_dismod_at.sh documentation has been simplified and its API has changed as follows:
The word
imageis no longer present in the image names.The image names have been changed as follows:
Old Name
New Name
dismod_at.basedismod_at.basedismod_at.mixed.build_typedismod_at.mixed.build_typedismod_at.image.build_typedismod_at.dismod_at.build_typeat_cascade.image.build_typedismod_at.at_cascade.build_type
01-03¶
On 2023 12-29 the dock_dismod_at.sh at_cascade_version did not advance to 2023.12.22 (it was 2023.11.20). This has been fixed.