I just fetched the most recent OG-USA files (v0.2.1), erased and rebuilt from scratch my ogusa-dev conda environement, and ran the run_og_usa.py run script. My baseline steady-state and transition path equilibria solved in just over 53 minutes. Then the reform started, got into the estimating new tax functions and stalled. Below is the terminal output of where the computation stalled. We need to get to the bottom of why this is happening.
cc: @jdebacker
Using reform policy tax parameters from /Users/richardevans/Docs/Economics/OSE/OG/OG-USA/examples/OG-USA-Example/OUTPUT_REFORM/TxFuncEst_policy.pkl
Using CPS
Using CPS
Using CPS
Using CPS
Using CPS
Using CPS
Using CPS
Calculator initial year = 2014
Calculator initial year = 2014
Calculator initial year = 2014
Calculator initial year = 2014
Calculator initial year = 2014
Calculator initial year = 2014
Calculator initial year = 2014
Year: 2025
Year: 2027
Year: 2032
Year: 2029
Year: 2030
Year: 2033
Year: 2034
Using CPS
Using CPS
Using CPS
Calculator initial year = 2014
Calculator initial year = 2014
Calculator initial year = 2014
Year: 2026
Year: 2031
Year: 2028
BW = 10 begin year = 2025 end year = 2034
/opt/anaconda3/envs/ogusa-dev/lib/python3.11/site-packages/distributed/client.py:3361: UserWarning: Sending large graph of size 224.31 MiB.
This may cause some slowdown.
Consider loading the data with Dask directly
or using futures or delayed objects to embed the data into the graph without repetition.
See also https://docs.dask.org/en/stable/best-practices.html#load-data-with-dask for more information.
warnings.warn(
Year= 2033 Age= 21
Year= 2033 Age= 22
Year= 2026 Age= 21
Year= 2031 Age= 21
Year= 2032 Age= 21
Year= 2027 Age= 21
Year= 2032 Age= 22
Year= 2033 Age= 23
Year= 2031 Age= 22
Year= 2026 Age= 22
...[LOTS OF YEARS AND AGES TAX FUNCTION COMPUTATION]...
Year= 2025 Age= 83
Insuff. sample size for age 83 in year 2025
Year= 2025 Age= 84
Insuff. sample size for age 84 in year 2025
Year= 2025 Age= 85
Linearly interpolate previous blank tax functions
Fill in all remaining old age tax functions.
Finished tax function loop through 10 years and 80 ages per year.
Tax function estimation time: 2.085 sec
ETR : 0 observations tagged as outliers.
MTRx : 2 observations tagged as outliers.
MTRx : After second round, 4 observations tagged as outliers (cumulative).
MTRy : 0 observations tagged as outliers.
Big S: 80
max age, min age: 100 21
making dir: /Users/richardevans/Docs/Economics/OSE/OG/OG-USA/examples/OG-USA-Example/OUTPUT_REFORM/SS
making dir: /Users/richardevans/Docs/Economics/OSE/OG/OG-USA/examples/OG-USA-Example/OUTPUT_REFORM/TPI
In runner, baseline is False
...[JUST HANGS HERE]...
I just fetched the most recent OG-USA files (v0.2.1), erased and rebuilt from scratch my
ogusa-dev
conda environement, and ran therun_og_usa.py
run script. My baseline steady-state and transition path equilibria solved in just over 53 minutes. Then the reform started, got into the estimating new tax functions and stalled. Below is the terminal output of where the computation stalled. We need to get to the bottom of why this is happening.cc: @jdebacker