Skip to content

RMG-Py v3.0.0 Release#1852

Merged
mliu49 merged 632 commits into
stablefrom
v3.0.0
Dec 17, 2019
Merged

RMG-Py v3.0.0 Release#1852
mliu49 merged 632 commits into
stablefrom
v3.0.0

Conversation

@mliu49

@mliu49 mliu49 commented Dec 16, 2019

Copy link
Copy Markdown
Contributor

This release represents a major milestone in RMG development and includes many backwards-incompatible changes, most notably Python 3 compatibility and major API changes. Users switching to RMG 3 will need to create new conda environments and update any scripts which access the API. We recommend using the futurize script from python-future for updating scripts for Python 3 and the provided rmg2to3.py script for updating scripts for RMG 3.

Python 3 #1724

Arkane

New features and other additions

Bug fixes

Other

Thanks to all contributors: ajocher, alongd, amarkpayne, cgrambow, dranasinghe, hwpang, kspieks, goldmanm, mazeau, mjohnson541, mliu49, oscarwumit, rwest, rgillis8, sarakha, sudoursa, xiaoruiDong, yunsiechung, zjburas

rwest and others added 30 commits September 29, 2019 18:22
Can we pickle and unpickle, reconstruct from __repr__,
and does the isIdenticalTo method work?

(The answer to the last two is currently No)
First, the Tmin and Tmax were being printed for Pmin and Pmax,
and the Pmin was being printed as the comment.

Second, the uncertainty was not being reported.
This method just checks that Tmin and Tmax are equal.
But if they were equal, it used to return False.

Now it returns False if one or other is None, or they're
different. Note, however, that because the Quantity.equals
method defaults to a 1% tolerance, that's 20K difference
needed at 2000K.
Fix a few bugs and add tests: KineticsModel and uncertainty handling.
The updated fingerprint approach resulted in different
fingerprint depending on atom order when the molecule contained
more than two elements aside from CHNOS.

This fixes the issue by sorting the elements alphabetically.
Fix inconsistent molecule fingerprint generation
The default value for the `force` argument changed from
False to True in matplotlib 3. However, we don't want to
force the backend change if we're using an ipython notebook.
added a try/except block around the clean build directories function. This should catch the error and provide a message instead of crashing the script.
Fix two calls to Exception.message, which no longer exists
…tiplebond_paths and find_adj_lone_pair_multiple_bond_delocalization_paths
In Python 2, negative numbers were treated as a single node
In Python 3, negative numbers are treated as a UnaryOp + a Num
Change shebang to point to python3 and remove encoding line
Ignore __pycache__ directories
Update name of Arkane.py script
We check that the user is using python 3 calling code in utilities.py
This file is not present though in the binary install. Thankfully, this
check is also not necessary in the binary install, as conda ensures the
correct python version for us. Therefore, we don't need to throw an error.
We currently don't support Windows
mliu49 and others added 27 commits December 12, 2019 10:27
The class was combined long ago with rmgpy.species.Species
Also fix some docstrings in MBSampledReactor
Use autodoc_default_options instead
I ran into issues related to negative predicted moments of inertia with this interpolant
This commit replaces that interpolant (the 2D case) with the LinearNDInterpolator
The old scheme using nquad only worked fast because the value of the integral was below the absolute tolerance, and it wasn't very accurate
after fixing the numerical issue nquad takes impractically long
since the integration is over a fit to data points we don't even know the function well enough for an nquad calculation to make sense
for this reason I've used simpsons rule for integration which is much faster and should be just as accurate
we still fit and use function evaluations for convenience
No need to set python version in meta.yaml
Update conda channels used for build
@codecov

codecov Bot commented Dec 16, 2019

Copy link
Copy Markdown

Codecov Report

Merging #1852 into stable will increase coverage by 2.54%.
The diff coverage is 66.43%.

Impacted file tree graph

@@            Coverage Diff             @@
##           stable    #1852      +/-   ##
==========================================
+ Coverage    41.7%   44.25%   +2.54%     
==========================================
  Files         176       83      -93     
  Lines       29341    21551    -7790     
  Branches     6049     5648     -401     
==========================================
- Hits        12237     9537    -2700     
+ Misses      16235    10944    -5291     
- Partials      869     1070     +201
Impacted Files Coverage Δ
arkane/kinetics.py 12.14% <ø> (ø)
arkane/sensitivity.py 10% <ø> (ø)
arkane/encorr/data.py 100% <ø> (ø)
arkane/ess/log.py 92.3% <100%> (ø)
arkane/ess/__init__.py 100% <100%> (ø)
arkane/encorr/__init__.py 100% <100%> (ø)
arkane/encorr/pbac.py 89.47% <100%> (ø)
arkane/encorr/mbac.py 16.94% <25%> (ø)
arkane/explorer.py 7.89% <5.45%> (ø)
arkane/ess/molpro.py 58.39% <53.96%> (ø)
... and 203 more

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 924c193...9548236. Read the comment docs.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.