core_analytics\local\target\binary Class Reference

Binary classifier target. More...

Inheritance diagram for core_analytics\local\target\binary:

Public Member Functions
	add_bulk_actions_js ()
	Adds the JS required to run the bulk actions.
	add_sample_data ($data)
	add_sample_data
	always_update_analysis_time ()
	Update the last analysis time on analysable processed or always.
	bulk_actions (array $predictions)
	Suggested bulk actions for a user.
	calculate ($sampleids, core_analytics\analysable $analysable, $starttime=false, $endtime=false)
	Calculates the target.
	can_use_timesplitting (\core_analytics\local\time_splitting\base $timesplitting)
	Can the provided time-splitting method be used on this target?.
	clear_sample_data ()
	clear_sample_data
	filter_out_invalid_samples (&$sampleids, core_analytics\analysable $analysable, $fortraining=true)
	Filters out invalid samples for training.
	generate_insight_notifications ($modelid, $samplecontexts, array $predictions=[])
	Generates insights notifications.
	get_analyser_class ()
	Returns the analyser class that should be used along with this target.
	get_calculation_outcome ($value, $ignoredsubtype=false)
	Is the calculated value a positive outcome of this target?
	get_display_value ($value, $ignoredsubtype=false)
	get_display_value
	get_id ()
	The class id is the calculable class full qualified class name.
	get_insight_body (\context $context, string $contextname, stdClass $user, moodle_url $insighturl)
	Returns the body message for an insight with multiple predictions.
	get_insight_body_for_prediction (\context $context, stdClass $user, core_analytics\prediction $prediction, array &$actions)
	Returns the body message for an insight for a single prediction.
	get_insight_context_url ($modelid, $context)
	URL to the insight.
	get_insight_subject (int $modelid, context $context)
	The insight notification subject.
	get_insights_users (\context $context)
	Returns the list of users that will receive insights notifications.
	ignored_predicted_classes ()
	Returns the predicted classes that will be ignored.
	is_linear ()
	is_linear
	is_valid_analysable (\core_analytics\analysable $analysable, $fortraining=true)
	Allows the target to verify that the analysable is a good candidate.
	is_valid_sample ($sampleid, core_analytics\analysable $analysable, $fortraining=true)
	Is this sample from the $analysable valid?
	link_insights_report ()
	Should the insights of this model be linked from reports?
	prediction_actions (\core_analytics\prediction $prediction, $includedetailsaction=false, $isinsightuser=false)
	Suggested actions for a user.
	prediction_callback ($modelid, $sampleid, $rangeindex, context $samplecontext, $prediction, $predictionscore)
	Callback to execute once a prediction has been returned from the predictions processor.
	save_calculation_info (\core_analytics\local\time_splitting\base $timesplitting, int $rangeindex)
	Stores in MUC the previously added data and it associates it to the provided $calculable.
	triggers_callback ($predictedvalue, $predictionscore)
	This method determines if a prediction is interesing for the model or not.

Static Public Member Functions
static	based_on_assumptions ()
	Based on facts (processed by machine learning backends) by default.
static	get_classes ()
	Returns the target discrete values.
static	get_name ()
	Returns a lang_string object representing the name for the indicator or target.
static	instance ()
	Returns an instance of the child class.
static	uses_insights ()
	Is this target generating insights?

Public Attributes
const	OUTCOME_NEGATIVE = 3
	Negative calculation outcome.
const	OUTCOME_NEUTRAL = 0
	Neutral calculation outcome.
const	OUTCOME_OK = 2
	Positive calculation outcome.
const	OUTCOME_VERY_NEGATIVE = 4
	Very negative calculation outcome.
const	OUTCOME_VERY_POSITIVE = 1
	Very positive calculation outcome.

Protected Member Functions
	add_shared_calculation_info (int $sampleid, array $info)
	Adds info related to the current calculation for later use when generating insights.
	calculate_sample ($sampleid, core_analytics\analysable $analysable, $starttime=false, $endtime=false)
	Calculates this target for the provided samples.
	classify_value ($value, $ranges)
	Classifies the provided value into the provided range according to the ranges predicates.
	get_time_range_weeks_number ($starttime, $endtime)
	Returns the number of weeks a time range contains.
	limit_value ($calculatedvalue)
	Limits the calculated value to the minimum and maximum values.
	min_prediction_score ()
	Defines a boundary to ignore predictions below the specified prediction score.
	retrieve ($elementname, $sampleid)
	Retrieve the specified element associated to $sampleid.

Static Protected Member Functions
static	classes_description ()
	classes_description
static	is_a_class ($class)
	Is the provided class one of this target valid classes?

Protected Attributes
core_analytics calculation_info null	$calculationinfo = null
array[]	$sampledata = array()

Detailed Description

Binary classifier target.

Copyright

2017 David Monllao

License

http://www.gnu.org/copyleft/gpl.html GNU GPL v3 or later

Member Function Documentation

add_bulk_actions_js()

core_analytics\local\target\base::add_bulk_actions_js ( )

inherited

Adds the JS required to run the bulk actions.

Reimplemented in core_course\analytics\target\course_enrolments.

add_sample_data()

core_analytics\calculable::add_sample_data ( $data )

inherited

add_sample_data

Parameters

array

$data

Return values

void

add_shared_calculation_info()

core_analytics\calculable::add_shared_calculation_info ( int $sampleid, array $info )

finalprotectedinherited

Adds info related to the current calculation for later use when generating insights.

Note that the data in $info array is reused across multiple samples, if you want to add data just for this sample you can use the sample id as key.

Please, note that you should be careful with how much data you add here as it can kill the server memory.

Parameters

int	$sampleid	The sample id this data is associated with
array	$info	The data. Indexed by an id unique across the site. E.g. an activity id.

Return values

null

always_update_analysis_time()

core_analytics\local\target\base::always_update_analysis_time ( )

inherited

Update the last analysis time on analysable processed or always.

If you overwrite this method to return false the last analysis time will only be recorded in DB when the element successfully analysed. You can safely return false for lightweight targets.

Return values

bool

Reimplemented in core_user\analytics\target\upcoming_activities_due.

based_on_assumptions()

static core_analytics\local\target\base::based_on_assumptions ( )

staticinherited

Based on facts (processed by machine learning backends) by default.

Return values

bool

Reimplemented in core_course\analytics\target\no_recent_accesses, core_course\analytics\target\no_teaching, and core_user\analytics\target\upcoming_activities_due.

bulk_actions()

core_analytics\local\target\base::bulk_actions ( array $predictions )

inherited

Suggested bulk actions for a user.

Parameters

core_analytics\prediction[]

$predictions

List of predictions suitable for the bulk actions to use.

Return values

core_analytics\bulk_action[]

The list of bulk actions.

Reimplemented in core_course\analytics\target\course_enrolments.

calculate()

core_analytics\local\target\base::calculate ( $sampleids, core_analytics\analysable $analysable, $starttime = false, $endtime = false )

inherited

Calculates the target.

Returns an array of values which size matches $sampleids size.

Rows with null values will be skipped as invalid by time splitting methods.

Parameters

array	$sampleids
core_analytics\analysable	$analysable
int	$starttime
int	$endtime

Return values

array

The format to follow is [userid] = scalar|null

calculate_sample()

core_analytics\local\target\base::calculate_sample ( $sampleid, core_analytics\analysable $analysable, $starttime = false, $endtime = false )

abstractprotectedinherited

Calculates this target for the provided samples.

In case there are no values to return or the provided sample is not applicable just return null.

Parameters

int	$sampleid
core_analytics\analysable	$analysable
int | false	$starttime	Limit calculations to start time
int | false	$endtime	Limit calculations to end time

Return values

float|null

can_use_timesplitting()

core_analytics\local\target\base::can_use_timesplitting ( \core_analytics\local\time_splitting\base $timesplitting )

abstractinherited

Can the provided time-splitting method be used on this target?.

Time-splitting methods not matching the target requirements will not be selectable by models based on this target.

Parameters

core_analytics\local\time_splitting\base

$timesplitting

Return values

bool

classes_description()

static core_analytics\local\target\binary::classes_description ( )

staticprotected

classes_description

Return values

string[]

Reimplemented from core_analytics\local\target\discrete.

Reimplemented in core_course\analytics\target\course_competencies, core_course\analytics\target\course_completion, core_course\analytics\target\course_dropout, core_course\analytics\target\course_gradetopass, core_course\analytics\target\no_teaching, and core_user\analytics\target\upcoming_activities_due.

classify_value()

core_analytics\calculable::classify_value ( $value, $ranges )

protectedinherited

Classifies the provided value into the provided range according to the ranges predicates.

Use:

• eq as 'equal'
• ne as 'not equal'
• lt as 'lower than'
• le as 'lower or equal than'
• gt as 'greater than'
• ge as 'greater or equal than'

Exceptions

coding_exception

Parameters

int | float	$value
array	$ranges	e.g. [ ['lt', 20], ['ge', 20] ]

Return values

float

clear_sample_data()

core_analytics\calculable::clear_sample_data ( )

inherited

clear_sample_data

Return values

void

filter_out_invalid_samples()

core_analytics\local\target\base::filter_out_invalid_samples ( & $sampleids, core_analytics\analysable $analysable, $fortraining = true )

inherited

Filters out invalid samples for training.

Parameters

int[]	$sampleids
core_analytics\analysable	$analysable
bool	$fortraining

Return values

void

generate_insight_notifications()

core_analytics\local\target\base::generate_insight_notifications ( $modelid, $samplecontexts, array $predictions = [] )

inherited

Generates insights notifications.

Parameters

int	$modelid
context[]	$samplecontexts
core_analytics\prediction[]	$predictions

Return values

void

get_analyser_class()

core_analytics\local\target\base::get_analyser_class ( )

abstractinherited

Returns the analyser class that should be used along with this target.

Return values

string

The full class name as a string

Reimplemented in core_course\analytics\target\course_enrolments, core_course\analytics\target\no_teaching, and core_user\analytics\target\upcoming_activities_due.

get_calculation_outcome()

core_analytics\local\target\binary::get_calculation_outcome	(		$value,
			$ignoredsubtype = false )

Is the calculated value a positive outcome of this target?

Parameters

string	$value
string	$ignoredsubtype

Return values

int

Reimplemented from core_analytics\local\target\discrete.

get_classes()

static core_analytics\local\target\binary::get_classes ( )

staticfinal

Returns the target discrete values.

Only useful for targets using discrete values, must be overwriten if it is the case.

Return values

array

Reimplemented from core_analytics\local\target\discrete.

get_display_value()

core_analytics\local\target\discrete::get_display_value ( $value, $ignoredsubtype = false )

inherited

get_display_value

Parameters

float	$value
string	$ignoredsubtype

Return values

string

Reimplemented from core_analytics\calculable.

get_id()

core_analytics\calculable::get_id ( )

inherited

The class id is the calculable class full qualified class name.

Return values

string

get_insight_body()

core_analytics\local\target\base::get_insight_body ( \context $context, string $contextname, stdClass $user, moodle_url $insighturl )

inherited

Returns the body message for an insight with multiple predictions.

This default method is executed when the analysable used by the model generates multiple insight for each analysable (one_sample_per_analysable === false)

Parameters

context	$context
string	$contextname
stdClass	$user
moodle_url	$insighturl

Return values

string[]

The plain text message and the HTML message

Reimplemented in core_course\analytics\target\course_enrolments, core_course\analytics\target\no_access_since_course_start, core_course\analytics\target\no_recent_accesses, and core_course\analytics\target\no_teaching.

get_insight_body_for_prediction()

core_analytics\local\target\base::get_insight_body_for_prediction ( \context $context, stdClass $user, core_analytics\prediction $prediction, array & $actions )

inherited

Returns the body message for an insight for a single prediction.

This default method is executed when the analysable used by the model generates one insight for each analysable (one_sample_per_analysable === true)

Parameters

context	$context
stdClass	$user
core_analytics\prediction	$prediction
core_analytics\action[]	$actions	Passed by reference to remove duplicate links to actions.

Return values

array

Plain text msg, HTML message and the main URL for this insight (you can return null if you are happy with the default insight URL calculated in prediction_info())

get_insight_context_url()

core_analytics\local\target\base::get_insight_context_url ( $modelid, $context )

inherited

URL to the insight.

Parameters

int	$modelid
context	$context

Return values

moodle_url

get_insight_subject()

core_analytics\local\target\base::get_insight_subject ( int $modelid, context $context )

inherited

The insight notification subject.

This is just a default message, you should overwrite it for a custom insight message.

Parameters

int	$modelid
context	$context

Return values

string

Reimplemented in core_course\analytics\target\course_enrolments, core_course\analytics\target\no_teaching, and core_user\analytics\target\upcoming_activities_due.

get_insights_users()

core_analytics\local\target\base::get_insights_users ( \context $context )

inherited

Returns the list of users that will receive insights notifications.

Feel free to overwrite if you need to but keep in mind that moodle/analytics:listinsights or moodle/analytics:listowninsights capability is required to access the list of insights.

Parameters

context

$context

Return values

array

get_name()

static core_analytics\calculable::get_name ( )

staticabstractinherited

Returns a lang_string object representing the name for the indicator or target.

Used as column identificator.

If there is a corresponding '_help' string this will be shown as well.

Return values

lang_string

Reimplemented in core\analytics\indicator\any_access_after_end, core\analytics\indicator\any_access_before_start, core\analytics\indicator\any_course_access, core\analytics\indicator\any_write_action, core\analytics\indicator\any_write_action_in_course, core\analytics\indicator\read_actions, core_course\analytics\indicator\activities_due, core_course\analytics\indicator\completion_enabled, core_course\analytics\indicator\no_student, core_course\analytics\indicator\no_teacher, core_course\analytics\indicator\potential_cognitive_depth, core_course\analytics\indicator\potential_social_breadth, core_course\analytics\target\course_competencies, core_course\analytics\target\course_completion, core_course\analytics\target\course_dropout, core_course\analytics\target\course_gradetopass, core_course\analytics\target\no_access_since_course_start, core_course\analytics\target\no_recent_accesses, core_course\analytics\target\no_teaching, core_user\analytics\indicator\user_profile_set, core_user\analytics\indicator\user_track_forums, core_user\analytics\target\upcoming_activities_due, mod_assign\analytics\indicator\cognitive_depth, mod_assign\analytics\indicator\social_breadth, mod_bigbluebuttonbn\analytics\indicator\cognitive_depth, mod_bigbluebuttonbn\analytics\indicator\social_breadth, mod_book\analytics\indicator\cognitive_depth, mod_book\analytics\indicator\social_breadth, mod_chat\analytics\indicator\cognitive_depth, mod_chat\analytics\indicator\social_breadth, mod_choice\analytics\indicator\cognitive_depth, mod_choice\analytics\indicator\social_breadth, mod_data\analytics\indicator\cognitive_depth, mod_data\analytics\indicator\social_breadth, mod_feedback\analytics\indicator\cognitive_depth, mod_feedback\analytics\indicator\social_breadth, mod_folder\analytics\indicator\cognitive_depth, mod_folder\analytics\indicator\social_breadth, mod_forum\analytics\indicator\cognitive_depth, mod_forum\analytics\indicator\social_breadth, mod_glossary\analytics\indicator\cognitive_depth, mod_glossary\analytics\indicator\social_breadth, mod_imscp\analytics\indicator\cognitive_depth, mod_imscp\analytics\indicator\social_breadth, mod_label\analytics\indicator\cognitive_depth, mod_label\analytics\indicator\social_breadth, mod_lesson\analytics\indicator\cognitive_depth, mod_lesson\analytics\indicator\social_breadth, mod_lti\analytics\indicator\cognitive_depth, mod_lti\analytics\indicator\social_breadth, mod_page\analytics\indicator\cognitive_depth, mod_page\analytics\indicator\social_breadth, mod_quiz\analytics\indicator\cognitive_depth, mod_quiz\analytics\indicator\social_breadth, mod_resource\analytics\indicator\cognitive_depth, mod_resource\analytics\indicator\social_breadth, mod_scorm\analytics\indicator\cognitive_depth, mod_scorm\analytics\indicator\social_breadth, mod_survey\analytics\indicator\cognitive_depth, mod_survey\analytics\indicator\social_breadth, mod_url\analytics\indicator\cognitive_depth, mod_url\analytics\indicator\social_breadth, mod_wiki\analytics\indicator\cognitive_depth, mod_wiki\analytics\indicator\social_breadth, mod_workshop\analytics\indicator\cognitive_depth, and mod_workshop\analytics\indicator\social_breadth.

get_time_range_weeks_number()

core_analytics\calculable::get_time_range_weeks_number ( $starttime, $endtime )

protectedinherited

Returns the number of weeks a time range contains.

Useful for calculations that depend on the time range duration. Note that it returns a float, rounding the float may lead to inaccurate results.

Parameters

int	$starttime
int	$endtime

Return values

float

ignored_predicted_classes()

core_analytics\local\target\binary::ignored_predicted_classes

(

)

Returns the predicted classes that will be ignored.

Return values

array

Reimplemented from core_analytics\local\target\discrete.

Reimplemented in core_user\analytics\target\upcoming_activities_due.

instance()

static core_analytics\local\target\base::instance ( )

staticinherited

Returns an instance of the child class.

Useful to reset cached data.

Return values

core_analytics\base\target

is_a_class()

static core_analytics\local\target\discrete::is_a_class ( $class )

staticprotectedinherited

Is the provided class one of this target valid classes?

Parameters

mixed

$class

Return values

bool

is_linear()

core_analytics\local\target\binary::is_linear

(

)

is_linear

Return values

bool

Reimplemented from core_analytics\local\target\discrete.

is_valid_analysable()

core_analytics\local\target\base::is_valid_analysable ( \core_analytics\analysable $analysable, $fortraining = true )

abstractinherited

Allows the target to verify that the analysable is a good candidate.

This method can be used as a quick way to discard invalid analysables. e.g. Imagine that your analysable don't have students and you need them.

Parameters

core_analytics\analysable	$analysable
bool	$fortraining

Return values

true|string

is_valid_sample()

core_analytics\local\target\base::is_valid_sample ( $sampleid, core_analytics\analysable $analysable, $fortraining = true )

abstractinherited

Is this sample from the $analysable valid?

Parameters

int	$sampleid
core_analytics\analysable	$analysable
bool	$fortraining

Return values

bool

limit_value()

core_analytics\calculable::limit_value ( $calculatedvalue )

protectedinherited

Limits the calculated value to the minimum and maximum values.

Parameters

float

$calculatedvalue

Return values

float|null

link_insights_report()

core_analytics\local\target\base::link_insights_report ( )

inherited

Should the insights of this model be linked from reports?

Return values

bool

Reimplemented in core_user\analytics\target\upcoming_activities_due.

min_prediction_score()

core_analytics\local\target\base::min_prediction_score ( )

protectedinherited

Defines a boundary to ignore predictions below the specified prediction score.

Value should go from 0 to 1.

Return values

float

prediction_actions()

core_analytics\local\target\base::prediction_actions ( \core_analytics\prediction $prediction, $includedetailsaction = false, $isinsightuser = false )

inherited

Suggested actions for a user.

Parameters

core_analytics\prediction	$prediction
bool	$includedetailsaction
bool	$isinsightuser	Force all the available actions to be returned as it the user who receives the insight is the one logged in.

Return values

core_analytics\prediction_action[]

prediction_callback()

core_analytics\local\target\base::prediction_callback ( $modelid, $sampleid, $rangeindex, context $samplecontext, $prediction, $predictionscore )

inherited

Callback to execute once a prediction has been returned from the predictions processor.

Note that the analytics_predictions db record is not yet inserted.

Parameters

int	$modelid
int	$sampleid
int	$rangeindex
context	$samplecontext
float | int	$prediction
float	$predictionscore

Return values

void

retrieve()

core_analytics\calculable::retrieve ( $elementname, $sampleid )

protectedinherited

Retrieve the specified element associated to $sampleid.

Parameters

string	$elementname
int	$sampleid

Return values

stdClass|false

An stdClass object or false if it can not be found.

save_calculation_info()

core_analytics\calculable::save_calculation_info ( \core_analytics\local\time_splitting\base $timesplitting, int $rangeindex )

finalinherited

Stores in MUC the previously added data and it associates it to the provided $calculable.

Flagged as final as we don't want people to extend this, it is likely to be moved to core_analytics\calculable

Parameters

core_analytics\local\time_splitting\base	$timesplitting
int	$rangeindex

Return values

null

triggers_callback()

core_analytics\local\target\discrete::triggers_callback ( $predictedvalue, $predictionscore )

inherited

This method determines if a prediction is interesing for the model or not.

This method internally calls ignored_predicted_classes to skip classes flagged by the target as not important for users.

Parameters

mixed	$predictedvalue
float	$predictionscore

Return values

bool

Reimplemented from core_analytics\local\target\base.

uses_insights()

static core_analytics\local\target\base::uses_insights ( )

staticinherited

Is this target generating insights?

Defaults to true.

Return values

bool

---

The documentation for this class was generated from the following file:

• analytics/classes/local/target/binary.php