Module: data_table¶
Module API¶
-
class
DatasetClassification
[source]¶ Bases:
asterism.core.machine_learning.data_sets.data_table.TableBase
Attributes
data
The numpy recarray storing the data Methods
add_column
(values_array, column_name[, dtype])Adds a column to the data, using numpy recfunctions filter_data
(mask)from_ascii_file
()from_fits_file
(input_file[, id_column_name, ...])set_data
(data[, target_col_num, ...])Parameters:
-
class
DatasetRegression
[source]¶ Bases:
asterism.core.machine_learning.data_sets.data_table.TableBase
Attributes
data
The numpy recarray storing the data Methods
add_column
(values_array, column_name[, dtype])Adds a column to the data, using numpy recfunctions filter_data
(mask)from_ascii_file
()from_fits_file
(input_file[, id_column_name, ...])set_data
(data[, target_col_num, ...])Parameters:
-
class
TableBase
(input_ndarray, id_column_name=None, id_column_num=None)[source]¶ Bases:
object
Parameters: input_ndarray : {array-like}, shape=[_data_N_rows,_data_N_cols]
id_column_name : str
name of the column storing positional/ordinal information
id_column_num : int
id of the column storing positional/ordinal information
Attributes
data
The numpy recarray storing the data _data_dtype (int) _data_array (int) _data_names (list) _data_N_cols (int) _data_N_rows (int) _data_original_entry_ID (int) Methods
add_column
(values_array, column_name[, dtype])Adds a column to the data, using numpy recfunctions filter_data
(mask)from_ascii_file
()from_fits_file
(input_file[, id_column_name, ...])set_data
(data[, target_col_num, ...])Parameters: -
add_column
(values_array, column_name, dtype=None)[source]¶ Adds a column to the data, using numpy recfunctions
Parameters: values_array : {array-like}, shape =[self._data_N_rows]
column_name : str
dtype :
-
data
¶ The numpy recarray storing the data
-
classmethod
from_fits_file
(input_file, id_column_name=None, id_column_num=None, fits_ext=0)[source]¶
-
set_data
(data, target_col_num=None, target_col_name=None, id_column_name=None, id_column_num=None)[source]¶ Parameters: data :
target_col_name : str
- name of the column storing the target information
Warning
this has been dismissed
target_col_num : int
- id of the column storing the target information
Warning
this has been dismissed
id_column_name : str
name of the column storing positional/ordinal information
id_column_num : int
id of the column storing positional/ordinal information
sets the following attributes:
- _data_dtype
- _data_array (dtype np.object)
- _data_names (list)
- _data_N_cols (int)
- _data_N_rows (int)
- _data_original_entry_ID 1d np.array np.int
Columns storing ID positional/ordinal information, or target (i.e. class labels for
classification, or target variables for regression are removed)
- Removes the column storing ID positional information (id_column_name/id_column_num if given)
- Removes the column storing target information (target_col_name/target_col_num if given)
Warning
this has been removed
The input data can be either a 2dim numpy array or, a numpy recarray.
In the former case, columns names are set to col_ plus the ordinal ID of the column, starting from zero
-