-
Notifications
You must be signed in to change notification settings - Fork 1.1k
Commit b0d29f5
Support for PV systems with multiple arrays (#1076)
* Initial implementation of pvsystem.Array class
Provides the three basic methods that depend on tilt and
azimuth and provides a container for other parameters
which may vary from array to array but do not vary within
a single array (racking, temperature model params,
albedo, strings, modules per string, etc.).
* Refactor PVSystem to use pvsystem.Array internally
Removes array-related fields and adds _array field. All internal references
to array-related attributes now refer to the the corresponding attributes of
the Array class.
It is likely some of the Array attributes will need to be exposed on the
PVSystem instances via @properties; however, any of these instances are not
covered by tests so it will be necessary to carefully work through the other
places where PVsystem is used (ModelChain and child classes) to see what
needs to be done.
* Use self._array.temperature_parameters in PVSystem
Use the temperature model parameters from the Array instance
inside PVSystem. Also properly initialize the parameters.
* Fix error passing surface_tilt in place of surface_azimuth
Mistake in invocation of Array.__init__() in PVSystem.__init__()
* Expose Array attributes as attributes of PVSystem
Uses `@property` to provide getters and setters for Array attributes
in order to mimic the original behavior of the PVSystem class.
* Pass attributes of PVSystem._array when constructing LocalizedPVSystem
* Fix over-indented lines
* Access Array attributes as attributes of the PVSystem
Since the Array attributes are all exposed via @propertys on PVSystem, we
can access them directly, reducing the number of changes needed to add the
Array class.
* Initialize Array.albedo correctly
* add ModelChainResults, plumb for airmass
* assign rest of ModelChain outputs to ModelChain.results
* first cut at ModelChain tests
* formatting
* more test development
* few more additions
* put dataclasses in pip section for py36
* Store a list of Arrays in PVSystem
Adds the @singleton_as_scalar decorator to make a PVSystem with a single
Array appear unchanged from the original PVSystem implementation.
* Wrap per-array inputs in a list
Adds a decorator (@list_or_scalar) used to specify which parameters
should be lists of the same length as PVSystem._arrays. The decorator
handles validation by transforming non-list values to singleton lists
then comparing to the length of the _arrays field on the PVSystem
instance passed as the first parameter.
Needs tests with multiple Arrays, but must wait until we have an API for
adding multiple Arrays to a PVSystem.
* Restore PVSystem._infer_cell_type()
This function is used in the test suite. It may not be needed otherwise,
but to move forward without needing to rewrite the tests I'm adding it
back to the PVSystem class as a wrapper around Array._infer_cell_type()
* Rename and document validation decorator
Renames `@list_or_scalar` to `@validate_against_arrays` and add a docstring
describing its behavior.
* Rename decorator for returning singleton list as a single value
Update docstring to make behavior more clear.
* Fix indentation and formatting errors
* Use nested def instead of assigning a lambda
Instead of assigning a lambda expression use an internal def to
define function for building the pvsyst_celltemp kwargs.
* Fix indentation in ModelChain._prepare_temperatures
* Add test for Array.__repr__()
* Pass list of Arrays to PVSystem.__init__
To construct a PVSystem instance with multiple arrays users can
pass a list of Array instances.
* Return tuple instead of list for multi-array PVSystem
Change the representation of PVSystem._array to a tuple as well.
* Fix indentation
* deprecate old ModelChain attributes
* remove solar_position = None from ModelChain.__init__
* Require tuple input in list_or_scalar decorator
After change to returning tuple as lists we need to update the
decortor to check for the correct type. It is tempting to test
for any Iterable here; however, since some of the parameters are
tuples of Series or DataFrame (which are themselves iterabel),
we need to be careful. I believe the best option is to be very
restrictive and expect input to be a tuple.
* Document meaning of `arrays` parameter to PVSystem.__init__
* Test multi-array PVSystem.scale_voltage_current_power()
* add modelchain test on pvsystem with 2 arrays
* Test multi-array PVSystem.get_aoi()
* Test multi-array PVSystem.get_irradiance()
Just test that PVSystem.get_irradiance returns a tuple that has the
irradiance for each array, test_PVSystem_get_irradiance() suffices to
test the the value returned by Array.get_irradiance() is correct.
* Test multi-array PVSystem.pvwatts_dc()
* Test input length validation on PVSystem.pvwatts_dc()
* Cover non-tuple iterable edge case
Per-Array input must be passed as a tuple, arbitrary iterable will raise
a ValueError even if the length is correct.
* Test multi-array PVSystem.get_iam()
* Fix whitespace in test_pvsystem.py
* Test PVSystem.get_iam() input length mismatch
* Test PVSystem.sapm with multiple arrays
* Test PVSystem.sapm_spectral_loss with multiple arrays
* Test PVSystem.sapm_effective_irradiance with multiple arrays
* Test PVSystem.sapm_celltemp() with multiple arrays
* Clean up whitespace
* Test PVSystem.first_solar_spectral_loss() with multiple arrays
* Test PVSystem.pvsyst_celltemp() with multiple arrays
Add two_array_system fixture to reduce duplicate setup code.
* Consolidate tests for PVSystem.xxxx_celltemp() functions
These all follow a similar pattern, so we can consolidate into two tests
patrametrized by a list of the celltemp functions to test.
* Clean up whitespace
* Fix comment in PVSystem.fuentes_celltemp()
Now we get the surface_tilt attribute form the Array, not PVSystem
* Test all PVSystem.calcparams_xxxx() functions with multiple arrays
* Add tests to cover PVSystem.albedo and PVSystem.surface_azimuth attributes
* Conditionally install dataclasses for tests
* formatting
* Remove TODOs
* Make decorators private
These decorators are not intended to be used anywhere else. Also rename
to avoid 'singleton' which is not being used in the standard "singleton
design pattern" sense.
* Update PVSystem.__repr__() to handle multiple Arrays
* Expect new format in tracking.SingleAxisTracker.__repr__() tests
Format has changed to support multi-array PVSystems.
* Update parameter names and doc for _validate_against_arrays()
* Replace validation decorator with a private method
The decorator adds too much complexity and room for error in the future.
A validation method that is called once on each parameter that needs to
be the same shape as the `self._arrays` tuple is much easier to read,
understand, and maintain. It also has the benefit of increased flexibility
since it allow validation to be triggered at any point, rather than before
the method is called.
* Remove unused `inspect` import in `pvsystem`
* Add blank line before nested defs
* lets see what breaks
* correct misspelled parameters
* infer methods, handle tuples in calculations
* inverter isn't on array
* fix iteration in ModelChain.effective_irradiance_model
* _tuple_from_dfs to function
* use isinstance for dataframe vs. tuple of df
* consolidate check for consistent parameters
* missing set()
* missed one instance
* Rename and document kwarg to prevent unwrapping return value
* Test ModelChain with multi-array system and no losses
Add a test that verifies the dc side of the ModelChain returns the correct
data for each Array with `aoi_model='no_loss'` and
`spectral_model='no_loss'`.
Test fails because `ModelChain.results.spectral_modifier` and
`ModelChain.results.aoi_modifier` are assigned a scalar value instead
of a tuple. It is very likely we will also see a failure at the
`ModelChain.losses_model()` call in `_run_from_effective_irradiance()`
for the same reason.
At this point we should (maybe) expect a failure at the call to
`ModelChain.ac_model()` since we have not implemented an inverter model
for multiple arrays yet.
* Initialize losses parameters corresponding to system._arrays
In the no_losses case, loss parameters are initialized to 1. This
commit initializes a tuple with the same number of elements as the
number of arrays in the system. In cases where losses are being
modeled, these fields are initialized to the result of a PVSystem
method call, meaning they have the correct type and shape. For cases
with no losses we need to take care to manually initialize them
correctly.
After this commit test_run_model_from_irradiance_arrays_no_loss()
fails at the call to PVSystem.pvwatts_ac() which is expected since
we have not implemented inverter models with multiple DC inputs yet.
* Fix whitespace
* Add PVSystem.num_arrays property
* Use self.system.num_arrays instead of len(self.system._arrays)
* Correct types of ModelChainResult fields
Declares a PerArray type which is either a single value or a tuple
of many values with the same type. The type for every field
that should have the same length as the number of arrays is
.Optional[PerArray[X]]
* Remove direct uses of ModelChain.system._arrays
Preserve encapsulation. Add an function _array_keys() that is
similar to _tuple_from_dfs in order to transparently handle the
different types returned by PVSystem attributes.
* Assign to self.results.effective_irradiance
Don't assign to the deprecated self.effective_irradiance
field internally.
* Test multi-array ModelChain with pvwatts losses
* Shorten line
* Test multi-array modelchain with DC models that call _singlediode
* Rewrite ModelChain._singlediode to support multi-array systems
Slightly different logic depending on the number of arrays. This could
be done a little more cleanly with the `unwrap=False` kwarg to
`calcparams_model_function` and `system.scale_voltage_current_power`.
That would give us a consistent type to work with, requiring only one
if statement to unwrap dc power ourselves if `system.num_arrays == 1`;
however, `mocker.spy` attempts to bind unwrap explicitly resulting in
a test failure where there should not be one, since unwrap is an extra,
non-explicit, parameter added by the `_unwrap_single_value` decorator.
* Add `name` attribute to `pvsystem.Array`
It may be useful to refer to Arrays by name in the future.
Adds an optional field to the Array class to support this.
* Allow per-array GHI/DHI/DNI input to PVSystem.get_irradiance()
* Test that the correct input is passed to each array
Validates the output of PVSystem.get_irradiance() by checking
against the output from Array.get_irradiance() with the expected
input for each Array.
* Fix whitespace in test_modelchain.py
* Support multiple weather input to PVSystem.xxxx_celltemp()
Support for system-wide or per-array values of air temperature and
wind speed.
* Fix whitespace in test_pvsystem.py
In tests:
- test_PVSystem_multi_array_celltemp_wind_too_long()
- test_PVSystem_multi_array_celltemp_temp_too_long()
* Allow unique weather input for each array
Updates ModelChain.prepare_inputs() to accept either a tuple of
DataFrames or a single DataFrame. If `weather` is a tuple each
DataFrame it contains must have the same index and must have all
of the required columns ('ghi', 'dhi', 'dni').
* Document that ModelChain.prepare_inputs() can take a tuple
Expand docstring to include new parameter type and note on the
exceptions that can be raised.
* Raise if length of weather is not same as system.num_arrays
Rather than raising the exception from deep in PVSystem we test the
length and raise immediately to make the error easier to understand
and debug for users.
* Add tests for ModelChain.run_model with multiple weather
* Add test for ModelChain._prepare_temperature with multiple weather
* Use itertools.starmap in _validate_weather_indices
Substantially more readable and concise.
* Update arrays in sapm_dc_snl_ac_system_same_arrays fixture
Use the sam array parameters as sapm_dc_snl_ac so we can use the same
expected output in multi-array tests.
* Rework _prepare_temperature tests
Add some assertions to verify that the cell temperature for both arrays
is not the same.
* ModelChain._prepare_temperature accepts multiple DataFrames
If multiple dataframes are passed in a tuple then _prepare_temperature
will look for 'cell_temperature' or 'module_temperature' in each data
frame. For data frames that do not have one of these columns the
ModelChain.temperature_model() is applied to calculate the cell
temperature from the weather data that has already been assigned in
other ModelChain fields.
* Fix indentation in ModelChain._get_cell_temperature()
* Support multiple weather DataFrames in all ModelChain.xxxx_temp()
Adds support for per-array weather input in
- ModelChain.pvsyst_temp()
- ModelChain.faiman_temp()
- ModelChain.fuentes_temp()
Support was added in ModelChain.sapm_temp() in a previous commit.
* Refactor ModelChain.xxxx_temp() functions
Abstract the pattern used in each function to reduce code duplication.
Should reduce the burden for adding new temperature models as well
as maintenance in the future.
* Accept multiple weather frames in ModelChain.complete_irradiance
Apply the same logic as in ModelChain.prepare_inputs to validate
the indices of the data frames. Iterate over them, adding missing
columns where necessary.
* ModelChain.complete_irradiance input same length as number of arrays
Document the new input types and exceptions.
* Validate per-array input to ModelChain.prepare_inputs_from_poa()
When passing POA irradiance directly to ModelChain, it must be passed
independently for each array. If the wrong number of POA dataframes
are provided by the user we raise a value error.
* Multiple inputs to ModelChain.run_model_from_effective_irradiance()
Multi-array systems require multiple effective irradiance DataFrames.
* Clean up whitespace
* Only validate indices if a tuple is passed
* Add test to cover undefined attribute
Because we hook `__getattr__` in order to deprecate the old
results attributes (e.g. `ModelChain.dc`) we need to test that
an AttributeError is raised when an unknown attribute is
requested.
* Verify all arrays have consistent parameters
Cliff wrote this. This commit just invokes the method and updates
it so it works with only one array.
* Don't test that all combinations of indices are equal
Because we are just comparing for equality we can infer from
a == b and b == c that a == c and don't have to explicitly
check every combination.
* Use ModelChain.results.dc instead of ModelChain.dc
ModelChain.dc will be deprecated when this is merged.
* Consolidate multi-input validation
Perform all validation for multi-array/multi-weather input
in a single method, `ModelChain._check_multiple_input()`.
* Add tests for multi-array module/string paramters
Tests PVSystem.modules_per_string and PVSystem.strings_per_inverter.
* Refactor all same index check
* Prevent instantiation of SingleAxisTracker with multiple arrays
Supporting tracking systems with multiple Arrays is a can of worms that
we don't want to open right now. Until support can be added we will
raise a ValueError if a user tries to instantiate a tracking system
with multiple arrays.
* Test that inconsistent array parameters causes error
If given a multi-array PVSystem where arrays have inconsistent module
or temperature model parameters a ValueError should be raised.
* Add test coverage for pvsystem.Array._infer_xxxx() methods
* Add pvsystem.Array to Classes section of api.rst
* Add PVSystem.sandia_multi() method
Adds a wrapper for the `pvlib.inverter.sandia_multi` inverter model
to PVSystem.
* Fix whitespace
* Add 'sandia_multi' as a valid key for ModelChain.ac_model
Adds the ModelChain.sandia_multi_inverter() method which calls
PVSystem.sandia_multi to set ModelChain.results.ac.
* Add sandia_multi_inverter to ModelChain.infer_ac_model()
Since no other inverter models currently support multiple dc inputs
we raise a value error whenever the system has more than one array and
is missing the inverter parameters required for sandia_multi_inverter.
* Update ModelChain tests to use sandia_multi inverter model
* Shorten long line
* Support multiple arrays in ModelChain.pvwatts_losses
* Decorate test_complete_irradiance_arrays with @requires_tables
* Pass explicit freq to DataFrame.shift() in poa_arrays test
* Pass explicit freq to DataFrame.shift() in ModelChain tests
In test_run_model_from_effective_irradiance_arrays_error() A different
index is generated by shifting the original input. Later versions of
pandas support freq='infer', but not the minimum supported version for
pvlib.
* Add description of multi-array PVSystem to pvsystem.rst
* Remove commented-out code in ModelChainResults
* Docstring edits
* Remove stray whitespace from _tuple_from_dfs() docstring
* Add docstring to ModelChain._check_consistent_params()
* Improvements in pvsystem.rst
Co-authored-by: Cliff Hansen <[email protected]>
* Docstring edits in pvsystem.py
Co-authored-by: Cliff Hansen <[email protected]>
* Shorten line in pvsystem.Array docstring
* Fix typo in ModelChain.complete_irradiance() docstring
Co-authored-by: Will Holmgren <[email protected]>
* Avoid use of `inplace=True` in ModelChain._singlediode()
* Make list of deprecated ModelChain attrs a class variable
* Include 'dataclasses' as a requirement in setup.py
Only included if python version is 3.6 since dataclasses is
part of the standard library for 3.7+.
* Remove dataclasses from CI environment for python 3.7+
* Replace modelchain._array_keys() with modelchain._common_keys()
_array_keys() just returned the keys of one specific array, what we
actually need is to return the set of keys that are common to every
dict in a tuple.
* Update ModelChain.__getattr__ deprecation message
Note that the deprecated attributes will be removed in 0.10, not
1.0.
* Expose PVSystem.arrays as a public attribute
This attribute holds the tuple of arrays in the PVSystem.
* Iterate over arrays in ModelChain._check_consistent_params()
Rather than iterating over individual parameters for each Array we
iterate over the arrays. Effectively the same, but much more readable.
* Refactor ModelChain.effective_irradiance_model()
Reduce code duplication by calling nested def for the single-array
case.
* rewrite pvsystem.rst
* Raise a TypeError in ModelChain._check_multiple_input()
When input must be a tuple, we should raise a TypeError, rather
than a ValueError.
* Note that indices assumed to be same in ModelChain._assign_times()
* Add comment in ModelChain.complete_irradiance()
Note why `ModelChain._assign_weather()` is not used here.
* Copy input before assigning in ModelChain.complete_irradiance()
To ensure that input data is not modified copy the data frames
that are passed to ModelChain.complete_irradiance() before assigning
`ModelChain.weather`.
* Document ModelChain.run_model() and run_model_from_poa() input type
Expand docstring to describe tuple input where each element specifies
the input for the corresponding Array.
* Don't allow SingleAxisTracker Array with non-None tilt/azimuth
If a list of one Array is given a value error is raised when it has
non-None tilt or azimuth. Rather than just assigning None we raise
an error so that the Array object itself is not modified (since it
could be in use by the user somewhere else). When only parameters
are passed (as opposed to an Array object) we override the values.
Since the Array is created internally and we know it is not in use
elsewhere this isn't a problem.
* Improve documentation of ValueErrors from ModelChain methods
Readability improvements by breaking up multiple reasons for a value
error into separate items.
* Separate ac model inference for multiple arrays
Adds an internal _infer_ac_model_multi() method to handle inference
when system.num_arrays > 1. This lets us raise a more usefule error
message in this case.
In addition several _xxxx_params() predicates are added to prevent
duplication of the specific tests on the inverter_params set in
_infer_ac_model_multi(). Seems like over-kill for now, but an
inverter.pvwatts_multi() model is in the works, so this will save
some work when integrating that.
* Update docstring for ModelChain.run_model_from_effective_irradiance()
Was referring to ModelChain.run_model_from() in the
"See also" section. That doesn't exist; changed to
ModelChain.run_model()
Co-authored-by: Will Holmgren <[email protected]>
* Expand error messages for missing array params in ModelChain
Note that all Arrays must have the correct/required model parameters
when an error is raised for missing parameters related to the inference
or assignment of the various models used in an instance of ModelChain.
* Remove ModelChain._check_consistent_params
Rework test to maintain coverage on inconsistent array parameters,
but expect the error to come from a different place. Mainly this means
that the error message has changed and more input is needed to the
constructor to make sure that execution goes all the way to the
assignment of the temperature model.
* Test ModelChain.infer_ac_model() error with one array
Make sure that an error is raised for a model chain with a single-
array system and invalid inverter_params.
* Capitalize 'Arrays' when referring to the "constituents" of a PVSystem
Apply consistent capitalization to make it clear when we are referring
to the idea of an array or an Array instance in a PVSystem.
* Return early from ModelChain._prepare_temperature()
If all Arrays have a cell_temperature or module_temperature given
in the input data then return early and don't try to compute
cell temperature based on weather. This supports calling
ModelChain.run_from_effective_irradiance() with only
'cell_temperature' and 'effective_irradiance' columns present in
the input.
* More test coverage of ModelChain.run_model_from_effective_irradiance()
* Note poa_global requirements for run_model_from_effective_irradiance()
Add not about when poa_global is required in the
ModelChain.run_model_from_effective_irradiance() docstring.
* Move ModelChain.tracking to ModelChain.results.tracking
Add tracking to deprecated attribute list.
* Improve pvsystem.Array docstring
Note that Array represents an array at a 'fixed orientation'
* Better documentation for ModelChain._check_multiple_input()
Provides a more thorough and informative description of the
`strict` parameter.
* Raise ValueError from ModelChain._prepare_temperature()
If the input data is missing cell_temperature for some Arrays
then the input must specify poa_global for _all_ Arrays.
* Update expected error message for run_model_from_effective_irradiance()
* coerce weather/data to tuple in run_model methods
* don't use decorator, _to_tuple in complete_irradiance
* adjust modelchain, add tests for [weather, weather], add test_run_model_from_poa_arrays
* Parameterize types in tests
Co-authored-by: Will Vining <[email protected]>
* add tuple/list type tests
* Use unwrap=True to simplify ModelChain._singldiode()
This lets us expect a consistent type from the `PVSystem` methods
regardless of the number of arrays. Because of that we can remove
the control-flow that made the function hard to read. This comes
at the cost of needing to unwrap the single value ourselves
before returning, but a single if statement at the end of the
method is much easier to reason about.
Because of the way mocks interact with wrapped functions we need to
move the mock from the object layer to the function layer, or it will
fail because the signature of the wrapped PVSystem methods does not
include the `unwrap` parameter.
* Add docstrings to helper functions
* Fix type annotations in ModelChainResults
Some fields had the wrong type (solar_position, airmass, ac, dc)
* Add index information to error message in ModelChain._verify()
Provide information to the caller about which input is missing
a column.
* fix copy, add input_type to wrong length tests
* Improve error message for inconsistent temperature model params
* Correct attribute name in ModelChain._set_celltemp docstring
self.results.cell_temp -> self.results.cell_temperature
* Remove unused ModelChainResult.array_ac field
* add coerce to prepare_inputs functions
* add _to_tuple back to run_model
* Coerce weather/data inputs to tuple in ModelChain entry points (#1)
* coerce weather/data to tuple in run_model methods
* add tuple/list type tests
* add input_type to wrong length tests
* add coerce to prepare_inputs functions
Co-authored-by: Will Vining <[email protected]>
* Add DatetimeIndex for test_prepare_inputs_multi_weather()
* Add whatsnew entries for 0.9.0
List of deprecated ModelChain attributes and brief summaries
of the main API enhancements.
* edits to pvsystem.rst
* fix ipython in pvsystem.rst
* edits to modelchain.py docstrings
* minor edits to whatsnew
* pvsystem.rst edits
* Allow Series input to modelchain._common_keys()
Return the intersection of the indices of the series. This
supports the use case shown in introtutorial.rst where
module parameters are a column selected from the a data frame.
* Update whatsnew v0.9.0 label
* Note dataclasses requirement in whatsnew/v0.9.0.rst
* Fix indentation in ModelChain.martin_ruiz_aoi_loss()
* Clean up Array.get_iam() docstring
Change PVSystem.module_parameters to Array.module_parameters.
Fix indentation of Raises section
* Change PVSystem to Array in Array.get_iam() error message
* Update links to other pvlib functions in Array.get_irradiance()
Links in docstring were broken. Use full name to fix.
* Compare function objects in ModelChain._get_cell_temperature()
Rather than using the name of the function we can compare
directly to the ModelChain.sapm_temp() bound method to check
whether the sapm temperature model is in use.
* merge master, edits to pvsystem.rst from review
* Reword multi-array SingleAxisTracker error message
remove the word "currently"
* Revert "Compare function objects in ModelChain._get_cell_temperature()"
This reverts commit 111ff32. Using `is` here was a mistake, and does not work.
* Update return types of PVSystem methods
Some methods now return tuples.
* Compare function objects in ModelChain._get_cell_temperature()
Rather than relying on the name of the function we can compare
function objects directly to determine if the sapm_temp
model has been selected by the user.
* Add project urls to setup.py for pypi page (#1119)
* Update setup.py
* Update setup.py
* add project_urls
* Update v0.8.1.rst
* Update v0.8.1.rst
* fix pvgis test (#1121)
* make clean: also delete generated, auto_examples, savefig (#1122)
* merge master, edits to pvsystem.rst from review
* Fix formatting in PVSystem.get_iam() docstring
* Pass model and kwargs to Array.get_irradiance()
Fixes bug where the `model` param and keyword arguments were
not passed when PVSystem.get_irradiance() called
Array.get_irradiance()
* Note which parameters of PVSystem methods accept tuples
Add note to parameter type in docstrings.
* Update Array docstring to describe modules "at same orientation"
* Document 0.25 as fallback value for Array.albedo
* Update defaults in Array docstring
* Rename and clean up duplicate PVSystem test
* Reword PVSystem.sandia_multi docstring
* Add notes on system-wide parameters to PVSystem method docstrings
* Add ModelChainResults to classes section of api.rst
* Expand ModelChain docstring to describe how models are applied
Make clear that the same model is applied to each array in the
system.
* Revert "Add ModelChainResults to classes section of api.rst"
This reverts commit 68617fa. The current sphinx configuration
can't handle this because ModelChainResult does not have
an explicit __init__ method. This results in an OSError during
the build when sphinx tries to make the source code pages.
* Simplify ModelChain.__getattr__()
Removes redundant look-up in self.__dict__
* Replace object with super() in ModelChain.__setattr__
* Use mc.results.ac instead of mc.ac in introtutorial.rst
* Replace deprecated ModelChain attributes in forecasts.rst
* fix pvsystem.rst examples
* Fix bug in pvsystem.rst examples
Forgot to pass module_parameters to PVSystem constructor.
* Add modelchain.ModelChainResult to Classes section of api.rst
Requires slight modification of sphinx/source/conf.py to address
getting line numbers for the ModelChainResult.__init__ method
which is not explicitly defined (and therefore does not have a line
number).
Made the T TypeVar private so it does not clutter the attribute list
for ModelChainResult.
Add docstring for the PerArray type.
* Update contributors list
Co-authored-by: Will Holmgren <[email protected]>
* Add ModelChain.solar_position to deprecated attrs in whatsnew
* Fix single array aoi example output
was printing the wrong aoi
* Add missing space in ValueError message from PVSystem.get_iam()
* Add Results to ModelChain section of api.rst
Co-authored-by: Will Vining <[email protected]>
Co-authored-by: Cliff Hansen <[email protected]>
Co-authored-by: Will Holmgren <[email protected]>
Co-authored-by: Aziz <[email protected]>
Co-authored-by: Kevin Anderson <[email protected]>1 parent ba7d753 commit b0d29f5Copy full SHA for b0d29f5
File tree
15 files changed
+3236
-574
lines changedFilter options
- ci
- docs/sphinx/source
- whatsnew
- pvlib
- tests
15 files changed
+3236
-574
lines changedci/requirements-py36-min.yml
Copy file name to clipboardExpand all lines: ci/requirements-py36-min.yml+1
| Original file line number | Diff line number | Diff line change | |
|---|---|---|---|
| |||
13 | 13 | | |
14 | 14 | | |
15 | 15 | | |
| 16 | + | |
16 | 17 | | |
17 | 18 | | |
18 | 19 | | |
| |||
+1
| Original file line number | Diff line number | Diff line change | |
|---|---|---|---|
| |||
27 | 27 | | |
28 | 28 | | |
29 | 29 | | |
| 30 | + | |
30 | 31 | | |
31 | 32 | | |
+8-1
| Original file line number | Diff line number | Diff line change | |
|---|---|---|---|
| |||
20 | 20 | | |
21 | 21 | | |
22 | 22 | | |
| 23 | + | |
23 | 24 | | |
24 | 25 | | |
25 | | - | |
| 26 | + | |
26 | 27 | | |
27 | 28 | | |
28 | 29 | | |
| |||
583 | 584 | | |
584 | 585 | | |
585 | 586 | | |
| 587 | + | |
| 588 | + | |
| 589 | + | |
| 590 | + | |
| 591 | + | |
| 592 | + | |
586 | 593 | | |
587 | 594 | | |
588 | 595 | | |
| |||
+4
| Original file line number | Diff line number | Diff line change | |
|---|---|---|---|
| |||
388 | 388 | | |
389 | 389 | | |
390 | 390 | | |
| 391 | + | |
| 392 | + | |
| 393 | + | |
| 394 | + | |
391 | 395 | | |
392 | 396 | | |
393 | 397 | | |
| |||
docs/sphinx/source/forecasts.rst
Copy file name to clipboardExpand all lines: docs/sphinx/source/forecasts.rst+3-3
| Original file line number | Diff line number | Diff line change | |
|---|---|---|---|
| |||
471 | 471 | | |
472 | 472 | | |
473 | 473 | | |
474 | | - | |
| 474 | + | |
475 | 475 | | |
476 | 476 | | |
477 | 477 | | |
| |||
482 | 482 | | |
483 | 483 | | |
484 | 484 | | |
485 | | - | |
| 485 | + | |
486 | 486 | | |
487 | 487 | | |
488 | 488 | | |
| |||
492 | 492 | | |
493 | 493 | | |
494 | 494 | | |
495 | | - | |
| 495 | + | |
496 | 496 | | |
497 | 497 | | |
498 | 498 | | |
| |||
docs/sphinx/source/introtutorial.rst
Copy file name to clipboardExpand all lines: docs/sphinx/source/introtutorial.rst+1-1
| Original file line number | Diff line number | Diff line change | |
|---|---|---|---|
| |||
165 | 165 | | |
166 | 166 | | |
167 | 167 | | |
168 | | - | |
| 168 | + | |
169 | 169 | | |
170 | 170 | | |
171 | 171 | | |
| |||
0 commit comments