lcc.stars_processing.tools package

Submodules

lcc.stars_processing.tools.params_estim module

class lcc.stars_processing.tools.params_estim.ParamsEstimator(searched, others, descriptors, deciders, tuned_params, split_ratio=0.5, static_params={}, **kwargs)[source]

Bases: object

Attributes

searched (list of Star objects) Searched stars
others (list of Star objects) Contamination stars
descriptors (list, iterable) Unconstructed descriptor objects
deciders (list, iterable) Decider instances
tuned_params (list of dicts) List of parameters to tune
static_params (dict) Constant values for descriptors and deciders

Methods

evaluate(combination)[source]
Parameters:

combination : dict

Dictionary of dictionaries - one per a descriptor.

EXAMPLE

{‘AbbeValue’: {‘bin’:10, .. }, .. }

Returns:

tuple

Stars filter, statistical values

evaluateCombinations()[source]

Evaluate all combination of the filter parameters

Returns:

list

Filters created from particular combinations

list

Statistical values of all combinations

list

Input parameters of all combinations

fit(score_func=None, opt='max', save_params={})[source]

Find the best combination of the filter parameters

Parameters:

score_func : function

Function which takes dict of statistical values and return a score

opt : str

Option for evaluating scores

“max” - Returns the highest score “min” - Returns the lowerest score

save_params : dict

Parameters for saving outputs. For each output there are some mandatory keys:

ROC plot:

“roc_plot_path” “roc_plot_name” “roc_plot_title” - optional

ROC data file:

“roc_data_path” “roc_data_name” “roc_data_delim” - optional

Statistical params of all combinations:

“stats_path” “stats_name” “stats_delim” - optional

Returns:

object

Filter created from the best parameters

dict

Statistical values of the best combination

dict

Input parameters of the best combination

saveOutput(save_params)[source]
Parameters:

save_params : dict

Parameters for saving outputs. For each output there are some mandatory keys:

ROC plot:

“roc_plot_path” “roc_plot_name” “roc_plot_title” - optional

ROC data file:

“roc_data_path” “roc_data_name” “roc_data_delim” - optional

Statistical params of all combinations:

“stats_path” “stats_name” “stats_delim” - optional

lcc.stars_processing.tools.stats_manager module

class lcc.stars_processing.tools.stats_manager.StatsManager(stats)[source]

Bases: object

Attributes

stats (list) List of dictionaries. They consists of statistical values. Or at least with “false_positive_rate” and “true_positive_rate” in order to work properly.

Methods

getROC()[source]

Get ROC curve

Returns:

list

List of fp values and tp values values

plotROC(save=False, title='ROC curve', path='.', file_name='roc_plot.png')[source]

Plot ROC and show it or save it

Parameters:

save : bool

If True plot is saved into the file

title : str

Title of the plot

path : str

Path to the output file location

file_name : str

Name of the file

Returns:

None

saveROCfile(path, file_name='roc_curve.dat', delim=None)[source]

Save ROC data into the file

Parameters:

path : str

Path to the output file location

file_name : str

Name of the file

delim : str

Delimiter of columns

Returns:

None

saveStats(path='.', file_name='stats.dat', delim=None, overwrite=True)[source]

Save stats file into the file

Parameters:

path : str

Path to the output file location

file_name : str

Name of the file

delim : str

Delimiter of columns

overwrite : bool

Overwrite file if it exists

Returns:

None

lcc.stars_processing.tools.visualization module

lcc.stars_processing.tools.visualization.plot1DProbabSpace(star_filter, plot_ranges, N, searched_coords=[], contaminatiom_coords=[])[source]

Plot probability space

Parameters:

star_filter : StarsFilter object

Trained stars filter

plot_ranges : iterable

Ranges (max/min) for all axis

N : int

Number of points per axis

searched_coords : list, iterable

List of coordinates of searched objects

contaminatiom_coords : list, iterable

List of coordinates of contamination objects

Returns:

tuple

x, y

lcc.stars_processing.tools.visualization.plot1DUnsupProbabSpace(coords, decider, opt, N)[source]
lcc.stars_processing.tools.visualization.plot2DProbabSpace(star_filter, plot_ranges, N, searched_coords=[], contaminatiom_coords=[])[source]

Plot probability space

Parameters:

star_filter : StarsFilter object

Trained stars filter

plot_ranges : iterable

Ranges (max/min) for all axis

N : int

Number of points per axis

searched_coords : list, iterable

List of coordinates of searched objects

contaminatiom_coords : list, iterable

List of coordinates of contamination objects

Returns:

tuple

x, y, Z

lcc.stars_processing.tools.visualization.plot2DUnsupProbabSpace(coords, decider, opt='show', N=50)[source]
lcc.stars_processing.tools.visualization.plotHist(searched_coo, cont_coo, labels=[], bins=None, save_path=None, file_name='hist.png')[source]

Plot histogram

Parameters:

searched_coo : iterable

Coordinates of searched objects to plot the histogram

cont_coo : iterable

Coordinates of contamination objects to plot the histogram

labels : list, tuple of str

Labels for axis

save_path : str, NoneType

Path to the folder where plots are saved if not None, else plots are showed immediately

bins : int, NoneType

Number of bins for histogram

file_name : str

Name of the plot file

Returns:

None

lcc.stars_processing.tools.visualization.plotProbabSpace(star_filter, plot_ranges=None, opt='show', path='.', file_name='params_space.png', N=400, title='Params space', x_lab='', y_lab='', searched_coords=[], contamination_coords=[], OVERLAY=0.6)[source]

Plot params space

Parameters:

star_filter : StarsFilter object

Trained stars filter object

plot_ranges : tuple, list

List of ranges. For example: [range(1,10), range(20,50)] - for 2D plot

opt : str

Option whether save/show/return

title : str

Title of the plot

path : str

Path to the output file location

file_name : str

Name of the file

OVERLAY : float

Percentage overlay of borders despite of data ranges

Returns:

None

lcc.stars_processing.tools.visualization.plotUnsupProbabSpace(coords, decider, opt='show', N=100)[source]

Module contents