Unsupervised clustering
See hint Hide hint
There are text input per every parameter of descriptors and deciders. You have to specify just one
value.
For evaluating content as python code wrapp the code into "`". For example:
`True` - bool value (not string)
`7*6` - integer (42)
`[("b_mag","v_mag"),("r_mag","i_mag")]` - list of tuples of strings
It is possible to select multiple descriptors and deciders.
NOTE that it raises error if loaded stars dont contain desired attribute (light curve, color index etc)
ColorIndexDescr
Filter star according their color indexes
Attributes
-----------
colors : list of strings
List of magnitudes which will be used. They are keys to color indexes
in star's object attribute 'more', where can be stored anything
pass_not_found : bool
If False stars without color index will be denied
raise_if_not : bool
If True it throws exception whenever a star has no color index
without_notfound : bool
If False coordinates of stars which have no color indexes will
be returned as well, but with None instead of coordinates (list of
values)
labels : list of strings
Labels of color-diagram axis
Modal Footer
VariogramSlopeDescr
This filter sorting stars according slopes of their variograms
Attributes
----------
days_per_bin : float
Rate between light curve dimension and days
absolute : bool
If True absolute value of slope is taken
Modal Footer
KurtosisDescr
KurtosisDescr describes stars by kurtosis
Attributes
----------
bins : int, NoneType
Dimension of reduced light curve. If it is `None` whole curve is taken.
absolute : bool
Absolute value of kurtosis is taken if it is `True`
Modal Footer
CurvesShapeDescr
This descriptor which compares light curves of inspected star
with the template in symbolic representation
Attributes
-----------
comp_stars : list
Template stars
days_per_bin : float
Ratio which decides about length of the word
alphabet_size : int
Range of of used letters
slide : bool
If True, words with different lengths are dynamically compared
by sliding shorter word thru longer
meth : str
Method key for calculating distance from comparative objects
average : take mean distance in each coordinate as
object coordinate
closest : take coordinate with closest distance as
object coordinate
best'n' : take best n scores of match, it can be integer or percentage float (0-1).
for example best10 takes 10 best matches, best0.5 takes 50 % best matches of the total
Modal Footer
AbbeValueDescr
AbbeValueDescr describes stars by Abbe values
Attributes
----------
bins : int
Dimension of reduced light curve from which Abbe value
is calculated
Modal Footer
CurveDescr
Attributes
----------
bins : int
Dimension of reduced light curve
height : int
Range of points in magnitude axis
red_dim : int, NoneType
If not None dimension is reduced by PCA into given size
Modal Footer
HistShapeDescr
This descriptor compares histograms of light curves of inspected star
with the template
Attributes
-----------
comp_stars : list
Template stars
bins : int
Length of result histogram
alphabet_size : int
Range of of used letters
slide : bool
If True, words with different lengths are dynamically compared
by sliding shorter word thru longer
slide : bool
If True, words with different lengths are dynamically compared
by sliding shorter word thru longer
meth : str
Method key for calculating distance from comparative objects
average : take mean distance in each coordinate as
object coordinate
closest : take coordinate with closest distance as
object coordinate
best'n' : take best n scores of match, it can be integer or percentage float (0-1).
for example best10 takes 10 best matches, best0.5 takes 50 % best matches of the total
Modal Footer
PropertyDescr
Descriptor which using star's attributes
Attributes
-----------
attribute_names : iterable, str
Keys of star's objects `more` attribute
For example:
`["pm_ra", "pm_de"]`
ifnot : str, NoneType
Value of coordinates which will be assigned if there is no
`attribute_name` value
Modal Footer
VariogramShapeDescr
This descriptor compares variograms of light curves of inspected star
with the template in symbolic representation.
Attributes
-----------
comp_stars : list
Template stars
bins : int
Number of bins
alphabet_size : int
Range of of used letters
slide : bool
If True, words with different lengths are dynamically compared
by sliding shorter word thru longer
meth : str
Method key for calculating distance from comparative objects
average : take mean distance in each coordinate as
object coordinate
closest : take coordinate with closest distance as
object coordinate
best'n' : take best n scores of match, it can be integer or percentage float (0-1).
for example best10 takes 10 best matches, best0.5 takes 50 % best matches of the total
Modal Footer
SkewnessDescr
SkewnessDescr describes stars by skewness
Attributes
----------
bins : int, NoneType
Dimension of reduced light curve. If it is `None` whole curve is taken.
absolute : bool
Absolute value of skewness is taken if it is `True`
Modal Footer
KMeansDecider
classdocs