backward integration bug

[mikelehu]

• Now, RKGaussLegendre.jl works well for backward integrations.
• I have added a new notebook called "/Examples/Simple Pendulum- Forward and Back integration.ipynb where I have tested the implementation

[codecov]

Codecov Report

Attention: 76 lines in your changes are missing coverage. Please review.

Comparison is base (38f1ca4) 25.44% compared to head (a84c095) 26.36%.

Files	Patch %	Lines
src/IRKGL16step_adaptive_par.jl	0.00%	24 Missing ⚠️
src/IRKGL16step_fixed_par.jl	0.00%	18 Missing ⚠️
src/IRKGL16step_adaptive_seq.jl	29.16%	17 Missing ⚠️
src/IRKGL16step_fixed_seq.jl	33.33%	12 Missing ⚠️
src/IRKGL16Solver.jl	76.19%	5 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##           master      #58      +/-   ##
==========================================
+ Coverage   25.44%   26.36%   +0.92%     
==========================================
  Files           8        8              
  Lines        1973     1995      +22     
==========================================
+ Hits          502      526      +24     
+ Misses       1471     1469       -2     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[ChrisRackauckas]

this file isn't needed

[ChrisRackauckas]

No test?

[mikelehu]

Hi,

Should I be worried about failed checks? Or will the pull request be merged into the base branch and made public in the next few days?

Thanks.

[ChrisRackauckas]

Those code coverage decreases are just saying that the new code isn't being tested.

[mikelehu]

I don't understand what you want to tell me. I have modified the implementation to resolve the issue raised by Andreas Varga.
issue 42

After confirming locally that the new deployment is correct, I made the pull request.
Do I delete the current pull request and create a new one?

[ChrisRackauckas]

You can add more commits to the same branch and it will register in this pull request. My point though is that the Jupyter notebook is not part of the test suite and it would be good to ensure that this feature gets a proper test. The only things that are tested are the files that are in the test/runtests.jl file. https://github.com/SciML/IRKGaussLegendre.jl/blob/master/test/runtests.jl I think if you paste the reverse time solve example into there with a quick test on it then this is good to go.

[mikelehu]

You are right, I agree that it is important to add a proper test for reverse time solve.
I have pasted an the example of the Jupyter Notebook in the runtests.jl file.
I hope it's enough, thanks for your help.