Open goerz opened 3 years ago
Check out this pull request on
See visual diffs & provide feedback on Jupyter Notebooks.
Powered by ReviewNB
Merging #92 (809d94e) into master (cfe6b10) will decrease coverage by
2.8%
. The diff coverage is45.4%
.
@@ Coverage Diff @@
## master #92 +/- ##
========================================
- Coverage 96.1% 93.2% -2.9%
========================================
Files 13 14 +1
Lines 1676 1765 +89
========================================
+ Hits 1611 1646 +35
- Misses 65 119 +54
Impacted Files | Coverage Δ | |
---|---|---|
src/krotov/__init__.py | 100.0% <ø> (ø) |
|
src/krotov/mu.py | 90.0% <33.3%> (-10.0%) |
:arrow_down: |
src/krotov/parametrization.py | 36.1% <36.1%> (ø) |
|
src/krotov/conversions.py | 95.6% <75.0%> (-2.5%) |
:arrow_down: |
src/krotov/optimize.py | 96.7% <75.0%> (-1.0%) |
:arrow_down: |
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 cfe6b10...809d94e. Read the comment docs.
This adds support for parametrized pulses by attaching an optional 'parametrization' attribute to control-functions and pulse arrays. The parametrization object stored in that attribute contains information for converting ϵ⟷u and calculating the derivative ∂ϵ/∂u. These are automatically taking into account when calculating the pulse updates (μ = (∂Ĥ/∂ϵ)⋅(∂ϵ/∂u) for an Ĥ that is linear in ϵ).
Thus, all internal storage is for the physical pulse values ϵ(t), and any conversion to u(t) is done on the fly as necessary. This ensures that everything stays 100% compatible with QuTiP, but adds a small overhead from the ϵ⟷u conversions compared to the "hacky" solution discussed in https://github.com/qucontrol/krotov/issues/23#issuecomment-850887732. The latter could still be used by an advanced user when performance is more critical than clarity.
Closes #23