All options for input

Here, all options for input are summarized. These input options correspond to the arguments of the pdc_sampler instance in the following Python code.

from aiphad import pdc_sampler

pdc  = pdc_sampler(page_type = "ternary_section",
                   input_data = "data.csv",
                   estimation = "LP",
                   gamma = 10,
                   sampling = "LC",
                   phase_id_option = {"FCC":0, "HCP":1, "BCC":2},
                   proposal = 3
                   )

The input of data has the following options.

A csv file is specified as variable input_data .
Input data is specified as arrays X and y .

In each case, the minimum required variables are as follows. 〇 indicates a required item, and × indicates an item that cannot be specified. Items marked with △ can be handled with either option.

variable	csv file as input data	arrays as input data	explanation
page_type	△	△	Enter the type of phase diagram corresponding to the web application version. If not, the figure cannot be outputed.
input_data	〇	×	Enter the file name when inputting a csv file.
estimation	〇	〇	Specify the LP or LS methods.
sampling	〇	〇	Specify LC, MS, EA, or RS methods．
phase_id_option	×	△	Specify phase name.
proposal	〇	〇	Specify the number of proposals.

The following are variables that can be specified as options.

variable	csv file as input data	arrays as input data	explanation
gamma	△	△	Specify \(\gamma\) of the rbf kernel used in the phase diagram estimation method.
parameter_constraint	△	△	Set to True to fix one variable from previous experimental condition. By default, False is selected.
prev_point	△	△	When fixing one variable, enter the previous proposal points as a list.
multi_method	△	△	Select either OU or NE when proposal is 2 or more. OU is selected by default.
NE_k	△	△	When NE is selected with multi_method, entier the value of NE_k. Candidate points closer than the NE_k nearest from the proposed point will not be proposed.