Skip to content

Commit

Permalink
iss #387 merged latest dev in this branch
Browse files Browse the repository at this point in the history
  • Loading branch information
@BiancaMorandi
BiancaMorandi committed May 30, 2023
2 parents 21d0a1e + 718f436 commit ed1e872
Show file tree
Hide file tree
Showing 11 changed files with 371 additions and 149 deletions.
1 change: 1 addition & 0 deletions CHANGELOG.md
View file
Original file line number Diff line number Diff line change
Expand Up @@ -41,6 +41,7 @@ _line_colors_, _legend_ and _figure_size_. (Issue #[349](https://github.com/brig
18. Fixed bug for `SpeedSort` where the `sector_predict` function was not interpolating data using two fit lines. (Issue #[377](https://github.com/brightwind-dev/brightwind/issues/377)).
19. Updated `_ColorPalette` to automatically update color_list, color_map, color_map_cyclical and adjusted lightness color variables when main colors (primary, secondary etc.) are changed. (Issue #[381](https://github.com/brightwind-dev/brightwind/issues/381)).
19. Allow `momm` function to derive a seasonal adjusted mean of monthly mean, if user sets `seasonal_adjustment` to true, and allow to apply a `coverage_threshold` (Issue #[298](https://github.com/brightwind-dev/brightwind/issues/298))
20. Updated `slice_data`, `offset_timestamps`, `_LoadBWPlatform.get_data` functions to use 'less than' data_to if provided as input. (Issue #[385](https://github.com/brightwind-dev/brightwind/issues/385))



Expand Down
18 changes: 13 additions & 5 deletions brightwind/analyse/analyse.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -377,7 +377,7 @@ def _mean_of_monthly_means_seasonal_adjusted(var_series):
return result


def momm(data, date_from: str = '', date_to: str = '', seasonal_adjustment=False, coverage_threshold=None):
def momm(data, date_from=None, date_to=None, seasonal_adjustment=False, coverage_threshold=None):
"""
Calculates and returns the mean of monthly mean speed. This accepts a DataFrame with timestamps as index column and
another column with wind speed. You can also specify date_from and date_to to calculate the mean of monthly
Expand All @@ -398,10 +398,18 @@ def momm(data, date_from: str = '', date_to: str = '', seasonal_adjustment=False
is not derived and the function will raise an error.
:param data: Pandas DataFrame or Series with timestamp as index and a column with wind speed
:type data: pd.DataFrame or pd.Series
:param date_from: Start date as string in format YYYY-MM-DD
:type data: pandas.DataFrame or pandas.Series
:param date_from: Start date as string in format YYYY-MM-DD or YYYY-MM-DD hh:mm. Start date is included in
the sliced data. If format of date_from is YYYY-MM-DD, then the first timestamp of the
date is used (e.g if date_from=2023-01-01 then 2023-01-01 00:00 is the first timestamp
of the sliced data). If date_from is not given then the sliced data are taken from the
first timestamp of the dataset.
:type: str
:param date_to: End date as string in format YYYY-MM-DD
:param date_to: End date as string in format YYYY-MM-DD or YYYY-MM-DD hh:mm. End date is not included in
the sliced data. If format date_to is YYYY-MM-DD, then the last timestamp of the
previous day is used (e.g if date_to=2023-02-01 then 2023-01-31 23:50 is the last
timestamp of the sliced data). If date_to is not given then the sliced data are taken up
to the last timestamp of the dataset.
:type: str
:param seasonal_adjustment: Optional, False by default. If True, returns the mean of monthly mean seasonally
adjusted
Expand All @@ -418,7 +426,7 @@ def momm(data, date_from: str = '', date_to: str = '', seasonal_adjustment=False
Default value is None, except when 'seasonal_adjustment'=True when it is 0.8.
:type coverage_threshold: int, float or None
:returns: Long term reference speed
:rtype: panda.Dataframe
:rtype: pandas.Dataframe
**Example usage**
::
Expand Down
64 changes: 57 additions & 7 deletions brightwind/analyse/plot.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -471,18 +471,24 @@ def _timeseries_subplot(x, y, x_label=None, y_label=None, x_limits=None, y_limit
return ax


def plot_timeseries(data, date_from='', date_to='', x_label=None, y_label=None, y_limits=None,
def plot_timeseries(data, date_from=None, date_to=None, x_label=None, y_label=None, y_limits=None,
x_tick_label_angle=25, line_colors=None, legend=True, figure_size=(15, 8)):
"""
Plot a timeseries of data.
:param data: Data in the form of a Pandas DataFrame/Series to plot.
:type data: pd.DataFrame, pd.Series
:param date_from: Start date used for plotting, if not specified the first timestamp of data is
considered. Should be in yyyy-mm-dd format
:param date_from: Start date as string in format YYYY-MM-DD or YYYY-MM-DD hh:mm. Start date is
included in the sliced data. If format of date_from is YYYY-MM-DD, then the first
timestamp of the date is used (e.g if date_from=2023-01-01 then 2023-01-01 00:00
is the first timestamp of the sliced data). If date_from is not given then the
sliced data are taken from the first timestamp of the dataset.
:type date_from: str
:param date_to: End date used for plotting, if not specified last timestamp of data is considered.
Should be in yyyy-mm-dd format
:param date_to: End date as string in format YYYY-MM-DD or YYYY-MM-DD hh:mm. End date is not
included in the sliced data. If format date_to is YYYY-MM-DD, then the last
timestamp of the previous day is used (e.g if date_to=2023-02-01 then
2023-01-31 23:50 is the last timestamp of the sliced data). If date_to is not given
then the sliced data are taken up to the last timestamp of the dataset.
:type date_to: str
:param x_label: Label for the x-axis. Default is None.
:type x_label: str, None
Expand Down Expand Up @@ -1695,6 +1701,49 @@ def plot_TI_by_sector(turbulence, wdir, ti):


def plot_shear_by_sector(scale_variable, wind_rose_data, calc_method='power_law'):
"""
Plot shear by directional sectors and wind rose.
:param scale_variable: Shear values by directional sectors derived with brightwind.Shear.BySector().
:type scale_variable: pandas.Series
:param wind_rose_data: Wind speed %frequency distribution by sectors, with wind direction sector as row indexes.
This distribution is derived using brightwind.dist_by_dir_sector() function.
:type wind_rose_data: pandas.Series
:param calc_method: Method to use for calculation, either 'power_law' (returns alpha) or 'log_law'
(returns the roughness coefficient).
:type calc_method: str
:return: Plots shear values by directional sectors & distribution of wind speed by directional bins.
**Example usage**
::
import brightwind as bw
data = bw.load_csv(bw.demo_datasets.demo_data)
alpha = pd.Series({'345.0-15.0': 0.216, '15.0-45.0': 0.196, '45.0-75.0': 0.170, '75.0-105.0': 0.182,
'105.0-135.0': 0.148, '135.0-165.0': 0.129, '165.0-195.0': 0.156, '195.0-225.0': 0.159,
'225.0-255.0': 0.160, '255.0-285.0': 0.169, '285.0-315.0': 0.187, '315.0-345.0': 0.188})
roughness = pd.Series({'345.0-15.0': 0.537, '15.0-45.0': 0.342, '45.0-75.0': 0.156, '75.0-105.0': 0.231,
'105.0-135.0': 0.223, '135.0-165.0': 0.124, '165.0-195.0': 0.135,
'195.0-225.0': 0.145, '225.0-255.0': 0.108, '255.0-285.0': 0.149,
'285.0-315.0': 0.263, '315.0-345.0': 0.275})
wind_rose_plot, wind_rose_dist = bw.analyse.analyse.dist_by_dir_sector(data.Spd80mS, data.Dir78mS,
direction_bin_array=None,
sectors=12,
direction_bin_labels=None,
return_data=True)
# Plots shear by directional sectors with calculation method as 'power law'.
bw.analyse.plot.plot_shear_by_sector(scale_variable=alpha, wind_rose_data=wind_rose_dist,
calc_method='power_law')
# Plots shear by directional sectors with calculation method as 'log law'.
bw.analyse.plot.plot_shear_by_sector(scale_variable=roughness, wind_rose_data=wind_rose_dist,
calc_method='log_law')
"""
result = wind_rose_data.copy(deep=False)
radians = np.radians(utils._get_dir_sector_mid_pts(scale_variable.index))
sectors = len(result)
Expand All @@ -1715,7 +1764,7 @@ def plot_shear_by_sector(scale_variable, wind_rose_data, calc_method='power_law'

scale_variable_y = np.append(scale_variable, scale_variable[0])
plot_x = np.append(radians, radians[0])
scale_to_fit = max(scale_variable) / max(result / 100)
scale_to_fit = max(scale_variable[np.isfinite(scale_variable)]) / max(result / 100)
wind_rose_r = (result / 100) * scale_to_fit
bin_edges = np.array(bin_edges)
width = pd.Series([], dtype='float64')
Expand All @@ -1732,7 +1781,8 @@ def plot_shear_by_sector(scale_variable, wind_rose_data, calc_method='power_law'
edgecolor=[COLOR_PALETTE.secondary for i in range(len(result))],
alpha=0.8, label='Wind_Directional_Frequency')

maxlevel = (max(scale_variable_y)) + max(scale_variable_y) * .1
maxlevel = (max(scale_variable_y[np.isfinite(scale_variable_y)])) + max(
scale_variable_y[np.isfinite(scale_variable_y)]) * .1
ax.set_thetagrids(radians * 180 / np.pi)
ax.plot(plot_x, scale_variable_y, color=COLOR_PALETTE.primary, linewidth=4, label=label)
ax.set_ylim(0, top=maxlevel)
Expand Down
99 changes: 52 additions & 47 deletions brightwind/load/load.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -1606,14 +1606,16 @@ def get_data(measurement_location_uuid, from_date=None, to_date=None):
"""
Retrieve measurement data from the brightwind platform and return it in a DataFrame with index as Timestamp.
:param measurement_location_uuid: The measurement location uuid.
:type measurement_location_uuid: str or uuid
:param from_date: Datetime representing the start of the measurement period you want.
:type from_date: datetime or str
:param to_date: Datetime representing the end of the measurement period you want.
:type to_date: datetime or str
:return: DataFrame with index as a timestamp.
:rtype: pd.DataFrame
:param measurement_location_uuid: The measurement location uuid.
:type measurement_location_uuid: str or uuid
:param from_date: Datetime representing the start of the measurement period you want
(included).
:type from_date: datetime or str
:param to_date: Datetime representing the end of the measurement period you want
(not included).
:type to_date: datetime or str
:return: DataFrame with index as a timestamp.
:rtype: pd.DataFrame
**Example usage**
::
Expand Down Expand Up @@ -1645,7 +1647,7 @@ def get_data(measurement_location_uuid, from_date=None, to_date=None):
if isinstance(from_date, str):
from_date = parse(from_date)
if isinstance(to_date, str):
to_date = parse(to_date)
to_date = parse(to_date) - datetime.timedelta(seconds=1)

response = requests.get(_LoadBWPlatform._base_url + '/api/resource-data-measurement-location', params={
'measurement_location_uuid': measurement_location_uuid,
Expand Down Expand Up @@ -1820,20 +1822,22 @@ def load_cleaning_file(filepath, date_from_col_name='Start', date_to_col_name='S
| Spd80m | 2018-10-23 12:30:00 | 2018-10-25 14:20:00
| Dir78m | 2018-12-23 02:40:00 |
:param filepath: File path of the file which contains the the list of sensor names along with the start and
end timestamps of the periods that are flagged.
:param filepath: File path of the file which contains the the list of sensor names along with the start
and end timestamps of the periods that are flagged.
:type filepath: str
:param date_from_col_name: The column name of the date_from or the start date of the period to be cleaned.
:type date_from_col_name: str, default 'Start'
:param date_to_col_name: The column name of the date_to or the end date of the period to be cleaned.
:type date_to_col_name: str, default 'Stop'
:param dayfirst: If your timestamp starts with the day first e.g. DD/MM/YYYY then set this to true. Pandas defaults
to reading 10/11/12 as 2012-10-11 (11-Oct-2012). If True, pandas parses dates with the day
first, eg 10/11/12 is parsed as 2012-11-10. More info on pandas.read_csv parameters.
:type dayfirst: bool, default False
:param kwargs: All the kwargs from pandas.read_csv can be passed to this function.
:return: A DataFrame where each row contains the sensor name and the start and end timestamps of the flagged data.
:rtype: pandas.DataFrame
:param date_from_col_name: The column name of the date_from or the start date of the period to be cleaned.
:type date_from_col_name: str, default 'Start'
:param date_to_col_name: The column name of the date_to or the end date of the period to be cleaned.
:type date_to_col_name: str, default 'Stop'
:param dayfirst: If your timestamp starts with the day first e.g. DD/MM/YYYY then set this to true.
Pandas defaults to reading 10/11/12 as 2012-10-11 (11-Oct-2012). If True, pandas parses
dates with the day first, eg 10/11/12 is parsed as 2012-11-10. More info on
pandas.read_csv parameters.
:type dayfirst: bool, default False
:param kwargs: All the kwargs from pandas.read_csv can be passed to this function.
:return: A DataFrame where each row contains the sensor name and the start and end timestamps of
the flagged data.
:rtype: pandas.DataFrame
**Example usage**
::
Expand Down Expand Up @@ -1870,32 +1874,33 @@ def apply_cleaning(data, cleaning_file_or_df, inplace=False, sensor_col_name='Se
| Spd80m | 2018-10-23 12:30:00 | 2018-10-25 14:20:00
| Dir78m | 2018-12-23 02:40:00 |
:param data: Data to be cleaned.
:type data: pandas.DataFrame
:param cleaning_file_or_df: File path of the csv file or a pandas DataFrame which contains the list of sensor
names along with the start and end timestamps of the periods that are flagged.
:type cleaning_file_or_df: str, pd.DataFrame
:param inplace: If 'inplace' is True, the original data, 'data', will be modified and and replaced with the cleaned
data. If 'inplace' is False, the original data will not be touched and instead a new object
containing the cleaned data is created. To store this cleaned data, please ensure it is assigned
to a new variable.
:param data: Data to be cleaned.
:type data: pandas.DataFrame
:param cleaning_file_or_df: File path of the csv file or a pandas DataFrame which contains the list of sensor
names along with the start and end timestamps of the periods that are flagged.
:type cleaning_file_or_df: str, pd.DataFrame
:param inplace: If 'inplace' is True, the original data, 'data', will be modified and and replaced
with the cleaned data. If 'inplace' is False, the original data will not be touched
and instead a new object containing the cleaned data is created. To store this
cleaned data, please ensure it is assigned to a new variable.
:type inplace: Boolean
:param sensor_col_name: The column name which contains the list of sensor names that have flagged periods.
:type sensor_col_name: str, default 'Sensor'
:param date_from_col_name: The column name of the date_from or the start date of the period to be cleaned.
:type date_from_col_name: str, default 'Start'
:param date_to_col_name: The column name of the date_to or the end date of the period to be cleaned.
:type date_to_col_name: str, default 'Stop'
:param all_sensors_descriptor: A text descriptor that represents ALL sensors in the DataFrame.
:type all_sensors_descriptor: str, default 'All'
:param replacement_text: Text used to replace the flagged data.
:type replacement_text: str, default 'NaN'
:param dayfirst: If your timestamp starts with the day first e.g. DD/MM/YYYY then set this to true. Pandas defaults
to reading 10/11/12 as 2012-10-11 (11-Oct-2012). If True, pandas parses dates with the day
first, eg 10/11/12 is parsed as 2012-11-10. More info on pandas.read_csv parameters.
:type dayfirst: bool, default False
:return: DataFrame with the flagged data removed.
:rtype: pandas.DataFrame
:param sensor_col_name: The column name which contains the list of sensor names that have flagged periods.
:type sensor_col_name: str, default 'Sensor'
:param date_from_col_name: The column name of the date_from or the start date of the period to be cleaned.
:type date_from_col_name: str, default 'Start'
:param date_to_col_name: The column name of the date_to or the end date of the period to be cleaned.
:type date_to_col_name: str, default 'Stop'
:param all_sensors_descriptor: A text descriptor that represents ALL sensors in the DataFrame.
:type all_sensors_descriptor: str, default 'All'
:param replacement_text: Text used to replace the flagged data.
:type replacement_text: str, default 'NaN'
:param dayfirst: If your timestamp starts with the day first e.g. DD/MM/YYYY then set this to true.
Pandas defaults to reading 10/11/12 as 2012-10-11 (11-Oct-2012). If True, pandas
parses dates with the day first, eg 10/11/12 is parsed as 2012-11-10.
More info on pandas.read_csv parameters.
:type dayfirst: bool, default False
:return: DataFrame with the flagged data removed.
:rtype: pandas.DataFrame
**Example usage**
::
Expand Down

0 comments on commit ed1e872

Please sign in to comment.