Bartdoekemeijer
commented
3 years ago I have refactored this PR and it is now ready to be merged into main. Important changes:
Major changes
energy_rationow accepts user-specified bin occurrences frequencies/probability distributions. This must be passed as an interpolant to theenergy_ratioclass. This need not necessarily be a normalized quantity: it will normalize the bin occurrences withinenergy_ratioafter interpolation, such that the total of bin occurrences/frequencies over all wind speeds for one single wind direction bin sums up to1.0.energy_ratio_suite.pynow can balance the bins between dataframes. This option defaults to True. What it does is that it calculates the bin_counts for all dataframes assigned to energy_ratio_suite and then determines the minimum bin count for each WD/WS bin among all dataframes. This is the frequency of occurrence for that particular bin. Once this balanced bin count has been established for all WD/WS bins, it is passed toenergy_ratioas a nearest-neighbor interpolant (see previous point). Bin balancing ensures we are comparing "apples to apples", so to speak.- Calculation of additional statistics concerning the energy ratio has been moved to
energy_ratio.py, rather than be calculated in the table generation function. This additional information is returned byenergy_ratioif the user specifiesreturn_detailed_output=True. Additional information includes: mean and standard deviation of the reference power, test power, wind direction, wind speed and optionally TI for every WD/WS bin. It also includes total energy produced (balanced and unbalanced), bin count/frequency of occurrence (balanced and unbalanced), energy ratios (balanced and unbalanced). - I have implemented it such that the table method will only work for one test_turbine at a time. Namely, the reference power can (and probably will) be different for every test_turbine, it seems better to me to generate one Excel file per test_turbine. @paulf81, I'm open to change my mind on this too. If so, let's get together and brainstorm.
energy_ratio_suitecurrently does not work for more than one test_turbine at a time.
Minor changes:
- The bin centers now starts at
wd_step/2.0(orwd_bin_width/2.0, whichever is smaller), and end at360.0. This means that the lower bound of the first bin is0.0, rather than starting the bin center at0.0and having the left bound be a negative number, and would then effectively only have data points in half of the first bin's bounds. Note that to include the case where the measurement is exactly0.0, we need to make sure the left bound of each bin is closed (wd_min <= wd < wd_max), rather than the right bound (wd_min < wd <= wd_max). I have changed this accordingly. - Add more docstrings to
energy_ratiofunctions. - Add TI/turbulence_intensity as a column to the energy ratio dataframes, if applicable.
As for validation:
- Energy ratios for demo/example case have not changed.
- Excel sheet produced by example case
energy_table_for_two_df.pyfrom Pauls original pull request matches that produced by the current code.
@paulf81 can you review this and let me know if any further changes are needed? If not, let me know and I'll merge it into the main repo. Thanks!
Really not ready to merge, just wanted to show the idea I've been building up in a notebook.
I think it's possible there is an error in the calculation, the energy ratios look too deep in some places and I don't think the directions make sense yet. But hopefully it makes an impression of this might work.