937f65f9c1
UPDATE: using **kwargs for build params
2021-03-25 20:39:35 -07:00
4b1a42d9ca
UPDATE: migrate all model.build_params to DictConfig
2021-03-25 20:38:52 -07:00
80b539eff6
UPDATE: notebook to use hydra config instead of yaml and rich.logging instead of manual prints
2021-03-25 20:38:25 -07:00
a6a97dfb5a
Merge branch 'main' into juan/ddm_evaluate
2021-03-16 11:23:07 -07:00
c4ae945ec2
Merge pull request #23 from microsoft/juan/ddm_fixes
...
[WIP] data validators by Juan
2021-03-16 11:22:14 -07:00
20c6c85a76
logger message for dir creation
2021-03-16 11:21:03 -07:00
df6f50f3b7
Removed User Story links from notebook
2021-03-16 00:51:53 -07:00
89759029a6
model validation notebook
...
Added notebook that makes use of the key features added to this branch
2021-03-16 00:29:01 -07:00
bb90d823b2
Bug fix
...
Commiting separately to ensure this is the correct fix
2021-03-16 00:27:16 -07:00
6ff69f1783
Add validation routines
...
Added sequential prediction from beginning to end;
Added sequential prediction that recovers original X value after certain iterations/steps;
Added evaluation functions for sequential analysis;
Consolidated evaluation functions on new python file
2021-03-16 00:24:54 -07:00
4378c51e9c
Merge branch 'juan/ddm_fixes' of github.com:microsoft/datadrivenmodel into juan/ddm_fixes
2021-03-15 03:59:14 -07:00
23c9a31fe0
Bug fixes
...
(1) creating folder for separate models in gboost_models.py & skmodels.py; (2) deepcopy in between fitting models in skmodels.py; (3) adding 'linear_model' to model_loader.py
2021-03-15 03:56:27 -07:00
88b13d8eb6
update: titles for distplots
2021-03-11 09:57:58 -08:00
2c77640a3b
Merge branch 'main' of github.com:microsoft/datadrivenmodel into main
2021-03-11 08:56:38 -08:00
37d6255245
Merge pull request #22 from microsoft/eda_notebook
...
Initital commit: EDA notebook
2021-03-11 08:56:26 -08:00
580d23f3a6
formatter
2021-03-11 08:55:06 -08:00
5045c7e08f
UPDATE: still wip; we should add an example where we actually randomized simconfigs during log-gathering
2021-03-11 08:51:24 -08:00
2d1ea5410d
Bug fixes
...
(1) creating folder for separate models in gboost_models.py & skmodels.py; (2) deepcopy in between fitting models in skmodels.py; (3) adding 'linear_model' to model_loader.py
2021-03-11 05:41:16 -08:00
b1c52eb770
Initital commit: EDA notebook
2021-02-28 21:08:19 -08:00
8e13f2da18
bugfix: evalutor
2021-02-25 14:45:49 -08:00
f7b673b6b3
REMOVE: old files
2021-02-25 14:35:23 -08:00
a0788b8d39
UPDATE: evaluate and evaluate margins methods
2021-02-25 14:14:57 -08:00
33be99fbee
UPDATE: evaluate and evaluate margins methods
2021-02-25 14:14:31 -08:00
c2433fc6a2
ADD: confs for hvac and moab
2021-02-18 14:23:42 -08:00
e517157e4a
UPDATE: add roc and auc visualizer
2021-02-18 14:14:11 -08:00
c353f86a86
UPDATE: evaluate and margin predictors
2021-02-18 14:02:46 -08:00
9266d5f714
Merge pull request #20 from microsoft/moab_example
...
Moab example
2021-01-21 21:08:56 +00:00
f8f295518f
more_conf
2021-01-19 20:44:29 -08:00
f49cf18516
more_conf_instructions
2021-01-19 15:11:37 -08:00
272eb4835a
moab_data_no_NaN
2021-01-19 14:48:57 -08:00
6a636c1347
UPDATE: moab example
2021-01-19 12:46:58 -08:00
5c4441db4a
rename ddm files
2021-01-19 09:40:30 -08:00
15dc9eff40
premerge
2021-01-19 09:37:43 -08:00
8954d96787
Merge pull request #19 from microsoft/alizaidi/refactor
...
WIP: Refactor with abstract classes to allow for different input / output pairs
2021-01-19 17:34:44 +00:00
7d18e348dd
update: use standard multioutputs loaders; update tests
2021-01-17 21:00:40 -08:00
2be904a6b5
remove: outputs dir
2021-01-16 22:45:25 -08:00
81f927c5bf
ADD: SVR.yml and tested
2021-01-15 18:00:09 -08:00
e133e814f1
UPDATE: if no augm columns then gracefully skip
2021-01-15 15:02:14 -08:00
aa81e93157
😰 update: final docs and passing off for users to test
2021-01-15 14:05:35 -08:00
e7eee6bd81
update ✊ : getting close to the end, docs and example complete
2021-01-15 14:03:31 -08:00
60cf508052
UPDATE ✊ sim_predictor complete with separate config and random test
2021-01-15 13:42:14 -08:00
a3bb51be10
:bugs: fix: fix saving when using single filename with gboost predictors
2021-01-15 12:32:46 -08:00
a7d37b7008
:bugs: fix: fix saving when using single filename with gboost predictors
2021-01-15 12:31:29 -08:00
9147cd243c
UPDATE: 📚 initial readme plus improved tests for sklearn and add max_rows arg
2021-01-15 11:48:54 -08:00
ca732bec49
ADD: configuration files for data conf with two examples
2021-01-15 10:11:21 -08:00
3f20261d7d
add: initial hydra integration for multiple configs
2021-01-15 00:16:44 -08:00
b0d7e43fc3
ADD: hydra dependencies to env yml
2021-01-15 00:14:44 -08:00
67c2d671cc
add episode and iter cols to base reader
2021-01-15 00:14:23 -08:00
5982153dc1
Merge branch 'alizaidi/refactor' of github.com:microsoft/datadrivenmodel into alizaidi/refactor
2021-01-14 21:42:16 -08:00
6035565840
premerge before hydra changes
2021-01-14 21:42:13 -08:00
Ali Zaidi
Juan Vergara
ucabmir
Brice Chung
HosseinHeris