apop_multinomial_probit.c File Reference

Functions

• apop_category_settings * apop_category_settings_alloc (apop_data *d, int source_column, char source_type)
• apop_category_settings * apop_category_settings_init (apop_category_settings in)
• void * apop_category_settings_copy (apop_category_settings *in)
• void apop_category_settings_free (apop_category_settings *in)
• apop_model * logit_estimate (apop_data *d, apop_model *m)

Variables

• apop_model apop_probit
• apop_model apop_logit
• apop_model apop_multinomial_probit

---

Detailed Description

---

Function Documentation

apop_category_settings* apop_category_settings_alloc	(	apop_data *	d,
		int	source_column,
		char	source_type
	)

Convert a column of input data into factors, for use with apop_probit, apop_logit, &c. Deprecated; use Apop_model_add_group and apop_category_settings_init.

Parameters:

	d	The input data set that you're probably about to run a regression on
	source_column	The number of the column to convert to factors. As usual, the vector is -1.
	source_type	't' = text; anything else ('d' is a good choice) is numeric data.

apop_category_settings* apop_category_settings_init

(

apop_category_settings

in

)

Convert a column of input data into factors, for use with apop_probit, apop_logit, &c.

You will probably use this with Apop_model_add_group, where you'll be specifying some of these inputs:

• .source_data The input data set that you're probably about to run a regression on
• .source_column The number of the column to convert to factors. As usual, the vector is -1.
• .source_type 't' = text; anything else ('d' is a good choice) is numeric data.

---

Variable Documentation

apop_model apop_logit

The Logit model. The first column of the data matrix this model expects a number indicating the preferred category; the remaining columns are values of the independent variables. Thus, the model will return N-1 columns of parameters, where N is the number of categories chosen.

apop_model apop_multinomial_probit

The Multinomial Probit model. The first column of the data matrix this model expects a number indicating the preferred category; the remaining columns are values of the independent variables. Thus, the model will return N-1 columns of parameters, where N is the number of categories chosen.

apop_model apop_probit

The Probit model. The first column of the data matrix this model expects is ones and zeros; the remaining columns are values of the independent variables. Thus, the model will return (data columns)-1 parameters.