Description

'Average Drift Across Versions' verses configure Mona to find segments, in which a metric's average differs significantly between a target and benchmark data sets - which are both defined by using a version field and defining which of its possible values is in each data set.
Note: Current implementation requires that versions are sorted alphabetically, it supports comparing the latest (alphabetically) version to any number of previous versions.

We then check if the difference between the average metric value in the target set to the average metric value in the benchmark set, normalized by the standard deviation (of the metric value. distribution in the united benchmark and target sets) is larger than a given minimal anomaly level.

{
  "stanzas": {
    "stanza_name": {
      "verses": [
        {
          "type": "AverageDriftAcrossVersions",
          "segment_by": [
            "company_id",
            "country"
          ],
          "metrics": [
            "confidence_score",
            "failed_classification"
          ],
          "min_segment_size_fraction": 0.05,
          "min_anomaly_level": 0.5,
          "version_field": "model_version",
          "num_versions_benchmark": 2,
          "order_versions_by": "LATEST_TO_RECEIVE_DATA"
        }
      ]
    }
  }
}

In this example we see an AverageDriftAcrossVersions verse which is configured to search for statistically significant changes in the average of "confidence_score" and "failed_classification" in any specific "company_id" or "country" (or any intersection of country and company), between the latest model version and the 2 versions prior to that. The latest model version is the latest version to receive
data.

We use "min_anomaly_level" to define that a drift occurs when the change in averages between the benchmark and target sets normalized by the standard deviation (of the metric value distribution in both benchmark and target sets) is larger than 0.5.

The "min_segment_size_fraction" param will filter out segments that are smaller than 5% of the data.

Basic Params

benchmark_set_period

Name	Description	Type	Default
benchmark_set_period	Time period for benchmark set. By default means the period just before the target set period. Expected format is "" where can be any positive integer, and options currently include: "d" (days), or "w" (weeks). e.g, "1d" means 1 day period.	TimePeriod

{ "benchmark_set_period": "50d" }

benchmark_set_period_type

Name	Description	Type	Default
benchmark_set_period_type	Sets the end time of the benchmark set period. Supports 'previous_to_target' (benchmark ends when target starts) and "latest" (both ends on the same date).	BenchmarkSetPeriodType	previous_to_target

{ "benchmark_set_period_type": "latest" }

metrics

Name	Description	Type	Default
metrics	Relevant metrics to search anomalies for in the verse, relevant only for types who search for anomalies in metrics behavior.	MetricsListType	[]

{ "metrics": [ "top_score", "delta_top_to_second_score" ] }

min_anomaly_level

Name	Description	Type	Default
min_anomaly_level	This parameter sets the threshold for the minimal anomaly level for which an insight will be generated. In standard mode - anomaly level is the diff between target and benchmark averages, normalized by overall STD. In no-STD-mode the anomaly threshold will be defined by "no_std_mode_min_diff_percent".	PositiveFloat	0.3

{ "min_anomaly_level": 0.6 }

min_segment_size

Name	Description	Type	Default
min_segment_size	Minimal segment size for the united benchmark+target segments.	PositiveInt	100

{ "min_segment_size": 100 }

min_segment_size_fraction

Name	Description	Type	Default
min_segment_size_fraction	Minimal segment size in fraction from baseline segment, which a segment must have in order to be considered in the search.	InclusiveFraction	0

{ "min_segment_size_fraction": 0.05 }

name

Name	Description	Type	Default
name	The name of the verse. If missing, this field defaults to "<stanza's name><verse's type>". Please note, a verse's name must be different from other verses in the same stanza. Therefore, in cases where there is more than one verse of the same type, using the default name ("<stanza's name><verse's type>") is not supported and custom names have to be provided for these verses.	None	None

{ "name": "confidence_outliers" }

num_versions_benchmark

Name	Description	Type	Default
num_versions_benchmark	Number of versions to set as benchmark, the target (last version in the alphabetical order) will be compared to its previous num_versions_benchmark versions. When set to -1, benchmark will be ALL previous versions.	NonNegativeOrMinusOneInt	1

{ "num_versions_benchmark": 2 }

order_versions_by

Name	Description	Type	Default
order_versions_by	The sorting manner of the versions. The last version in the chosen ordering manner will be the target and the previous num_versions_benchmark versions on this order will be the benchmark. order_versions_by options: "ALPHABETICAL": The target will be the last version in the alphabetical order. "LATEST_TO_RECEIVE_DATA": The target will be the latest version according to the last time data was received for it. "LATEST_TO_RECEIVE_FIRST_DATA": The target will be the latest version according to the first time data was received for it.	OrderVersionsBy	ALPHABETICAL

{ "order_versions_by": "LATEST_TO_RECEIVE_DATA" }

segment_by

Name	Description	Type	Default
segment_by	The dimensions to use to segment the data in order to search for anomalies. This list must be a sublist of all arc class' dimensions. Avoiding insights generation on a specific value of a segmentation field can be done by using the "avoid_values" key in the segmentation json object, as seen in the example.	SegmentationsListType	[]

{ "segment_by": [ "city", "bot_id", {"name": "provider-code", "avoid_values": ["zoom"]}, {"name": "selected-language", "avoid_values": ["eng", "spa"]} ] }

target_set_period

Name	Description	Type	Default
target_set_period	Time period for the target set, ending on the day of the latest available data. Format detailed in common/util.py's get_time_period_for_string.	TimePeriod

{ "target_set_period": "1w" }

time_resolution

Name	Description	Type	Default
time_resolution	The time resolution of time series charts for insights of this verse. NOTE: this parameter doesn't affect the behavior of the insights finding mechanism, but only the data representation in the output.	VisualTimeResolution	1d

{ "time_resolution": "1w" }

trend_directions

Name	Description	Type	Default
trend_directions	A list of allowed anomalies trends directions - either 'asc' for ascending (anomalies in which the found value is LARGER THAN the relevant benchmark), or 'desc' for descending (anomalies in which the found value is SMALLER THAN the relevant benchmark).	TrendDirectionsType	['desc', 'asc']

{ "trend_directions": [ "asc" ] }

version_field

Name	Description	Type	Default
version_field	The name of the field containing the versions to be compared (required).	str	None

{ "version_field": "model_version" }

Advanced Misc Params

avoid_same_field_for_segment_and_metric

Name	Description	Type	Default
avoid_same_field_for_segment_and_metric	If True, insights would not be created for segments based on the same field as the given metric.	bool	True

{ "avoid_same_field_for_segment_and_metric": false }

cadence

Name	Description	Type	Default
cadence	The cadence for evaluation of this verse. Only the following cadences are valid: Minutes: 1m, 5m, 10m, 15m, 20m, 30m. Hours: 1h, 2h, 3h, 4h, 6h, 8h, 12h. Days: 1d, 2d, 3d, 4d, 5d, 6d. Weeks: 1w, 2w, 3w, 4w, 5w.	CadenceType	1d

{ "cadence": "6h" }

create_extra_adjacent_signals

Name	Description	Type	Default
create_extra_adjacent_signals	If set to true (default), will cause Mona to create new signals from existing signals with adjacent numeric segments. So if there are two signals defined on 1 <= x < 2 and 2 <= x < 3 - Mona will automatically create a new signal with 1 <= x < 3. This will allow the later clustering algorithm to create an insight with the most relevant segment for its main signal.	bool	True

{ "create_extra_adjacent_signals": false }

disabled

Name	Description	Type	Default
disabled	If set to True - this verse won't be used when searching for new insights.	bool	False

{ "disabled": true }

expire_after

Name	Description	Type	Default
expire_after	Insights detected by this verse will continue to be considered active for at least this amount of time after the last time they were detected.	TimePeriod	3d

{ "expire_after": "2d" }

Advanced Score Calculation Params

no_std_mode_min_segment_size

Name	Description	Type	Default
no_std_mode_min_segment_size	Minimum segment size to start using no-STD mode. -1 means don't use this method at all. 0 means only use no_std_mode. If this is positive, smaller segments will use the std-aware mode.	NonNegativeOrMinusOneInt	-1

{ "no_std_mode_min_segment_size": 0 }

no_std_mode_reference_values

Name	Description	Type	Default
no_std_mode_reference_values	In no-STD mode, use these values as reference to compare both benchmark and target averages.	frozenset	[0]

{ "no_std_mode_reference_values": [ 0, 1 ] }

score_anomaly_level_exponent

Name	Description	Type	Default
score_anomaly_level_exponent	An exponent to put on the anomaly level in the score after multiplying it by the given multiplier.	float	1.0

{ "score_anomaly_level_exponent": 0.5 }

score_anomaly_level_multiplier

Name	Description	Type	Default
score_anomaly_level_multiplier	Multiplier for an anomaly level to use before using the exponent.	float	1.0

{ "score_anomaly_level_multiplier": 1.2 }

score_segment_size_exponent

Name	Description	Type	Default
score_segment_size_exponent	An exponent to put on the segment's size (or relative size) in the combined score. If score_segment_size_log_base is not 0, the exponent will be applied before the logarithm will.	float	0.5

{ "score_segment_size_exponent": 1.5 }

score_segment_size_log_base

Name	Description	Type	Default
score_segment_size_log_base	Changes the log base used for the segment's size (or relative size) in the combined score, or remove the log altogether by setting 0 here. Unless its 0 this value must be larger than 1.	float	0.0

{ "score_segment_size_log_base": 5 }

score_use_segment_absolute_size

Name	Description	Type	Default
score_use_segment_absolute_size	If true, use the segment absolute size in the combined score, otherwise use the segment's size relative to its baseline (fraction).	bool	True

{ "score_use_segment_absolute_size": false }

Anomaly Thresholds Params

epsilon

Name	Description	Type	Default
epsilon	Minimal absolute difference between benchmark and value.	NonNegativeFloat	0.01

{ "epsilon": 0.5 }

min_anomaly_level

Name	Description	Type	Default
min_anomaly_level	This parameter sets the threshold for the minimal anomaly level for which an insight will be generated. In standard mode - anomaly level is the diff between target and benchmark averages, normalized by overall STD. In no-STD-mode the anomaly threshold will be defined by "no_std_mode_min_diff_percent".	PositiveFloat	0.3

{ "min_anomaly_level": 0.6 }

min_score

Name	Description	Type	Default
min_score	The minimal score for a signal to be considered as an anomaly.	float	0.0

{ "min_score": 4 }

no_std_mode_min_diff_percent

Name	Description	Type	Default
no_std_mode_min_diff_percent	In no-STD mode, acts as the anomaly threshold and is the minimal difference required in percents, calculated according to the proximity to the closest reference point, as configured in the param no_std_mode_reference_values.	PositiveFloat	25.0

{ "no_std_mode_min_diff_percent": 30.0 }

Data Filtering Params

avoid_segmenting_on_missing

Name	Description	Type	Default
avoid_segmenting_on_missing	When true, insights will not be generated for segments which are (partially or fully) defined by a missing field.	bool	False

{ "avoid_segmenting_on_missing": true }

baseline_segment

Name	Description	Type	Default
baseline_segment	The baseline segment of this verse. This segment defines "the world" as far as this verse is concerned. Only data from this segment will be considered when finding insights.	Segment	{}

{ "baseline_segment": { "model_version": [ { "value": "V1" } ] } }

benchmark_baseline_segment

Name	Description	Type	Default
benchmark_baseline_segment	Benchmark baseline segment. This segment is intersected with any data we search for in the benchmark segments.	Segment	{}

{ "benchmark_baseline_segment": { "model_version": [ { "value": "V2" } ] } }

enhance_exclude_segments

Name	Description	Type	Default
enhance_exclude_segments	If True, when exclude segments are added to any level of configuration (either in the verse, the stanza or the stanzas_global_defaults) they are ADDED to the excluded segments of higher level defaults, if exists any. For example, if we have in stanzas_global_default a single excluded segment of {dimensionA: MISSING}, and the stanza (or verse) has a single excluded segment of {dimensionB: 0}, then if enhance_exclude_segments is True, the excluded segments will include both {dimensionA: MISSING} and {dimensionB: 0} and will filter either one. Otherwise (if enhance_exclude_segments is False) it will be overridden to just the one segment in the verse {dimensionB: 0}.	bool	False

{ "enhance_exclude_segments": true }

exclude_segments

Name	Description	Type	Default
exclude_segments	Segments to exclude in the baseline of this verse. Each data we search for will not include these segments - both tested segments as well as any benchmarks used to find the anomalies. Notice that whether or not this param will override definitions for exclude_segments in other levels is decided by enhance_exclude_segments.	SegmentsListType	[]

{ "exclude_segments": [ { "text_length": [ { "min_value": 0, "max_value": 100 } ] } ] }

target_baseline_segment

Name	Description	Type	Default
target_baseline_segment	Target baseline segment. This segment is intersected with any data we search for in the tested segments.	Segment	{}

{ "target_baseline_segment": { "model_version": [ { "value": "V1" } ] } }

Related Anomalies Params

avoid_related_anomalies_for

Name	Description	Type	Default
avoid_related_anomalies_for	A list of fields to avoid checking for correlated anomalies to the main anomaly in a generated insight. See "find_related_anomalies_for" for further details.	MetricsListType	[]

{ "avoid_related_anomalies_for": ["delta_top_to_second_score"] }

find_related_anomalies_for

Name	Description	Type	Default
find_related_anomalies_for	A list of fields to check for correlated anomalies to the main anomaly in a generated insight. These correlated anomalies might help with understanding the possible cause of an insight. Leave empty to search in all fields.	MetricsListType	[]

{ "find_related_anomalies_for": ["sentiment_score", "confidence_interval"] }

related_anomalies_min_correlation

Name	Description	Type	Default
related_anomalies_min_correlation	Minimal Pearson correlation between the metric on which an anomaly was found and another metric with an anomaly on the same segment, below which Mona will not use the other metric as a related anomaly.	NonNegativeFloat	0.3

{ "related_anomalies_min_correlation": 0.5 }

Segmentation Params

always_segment_baseline_by

Name	Description	Type	Default
always_segment_baseline_by	A list of dimensions to always segment the baseline segment by. This is useful when separating the world to completely unrelated parts - e.g., in a case where you have a different model developed for each customer and there's no need to look for insights across different customers. Avoiding insights generation on a specific value of a segmentation field can be done by using the "avoid_values" key in the segmentation json object, as seen in the example.	SegmentationsListType	[]

{ "always_segment_baseline_by": [ "country", {"name": "city", "avoid_values": ["Tel Aviv"]}, ] }

avoid_segmenting_on_missing

Name	Description	Type	Default
avoid_segmenting_on_missing	When true, insights will not be generated for segments which are (partially or fully) defined by a missing field.	bool	False

{ "avoid_segmenting_on_missing": true }

max_segment_baseline_by_depth

Name	Description	Type	Default
max_segment_baseline_by_depth	The maximum number of fields Mona should combine for segmenting the baseline (if "segment_baseline_by" fields given).	PositiveInt	2

{ "max_segment_baseline_by_depth": 3 }

max_segment_by_depth

Name	Description	Type	Default
max_segment_by_depth	The maximum number of fields Mona should combine to create sub-segments to search in. Baseline segment fields and parent fields are "free", and are not counted for depth. Notice this parameter has a exponential effect on the performance and should be kept within SLAs.	PositiveInt	2

{ "max_segment_by_depth": 3 }

min_segment_baseline_by_depth

Name	Description	Type	Default
min_segment_baseline_by_depth	The minimum number of fields Mona should combine for segmenting the baseline (if "segment_baseline_by" fields given).	NonNegativeInt	0

{ "min_segment_baseline_by_depth": 1 }

min_segment_by_depth

Name	Description	Type	Default
min_segment_by_depth	The minimum number of fields Mona should combine to create sub-segments to search in.	NonNegativeInt	0

{ "min_segment_by_depth": 1 }

segment_baseline_by

Name	Description	Type	Default
segment_baseline_by	A list of dimensions to potentially segment the baseline segment by. Avoiding insights generation on a specific value of a segmentation field can be done by using the "avoid_values" key in the segmentation json object.	SegmentationsListType	[]

{ "segment_baseline_by": [ "model_version" ] }

segment_by

Name	Description	Type	Default
segment_by	The dimensions to use to segment the data in order to search for anomalies. This list must be a sublist of all arc class' dimensions. Avoiding insights generation on a specific value of a segmentation field can be done by using the "avoid_values" key in the segmentation json object, as seen in the example.	SegmentationsListType	[]

{ "segment_by": [ "city", "bot_id", {"name": "provider-code", "avoid_values": ["zoom"]}, {"name": "selected-language", "avoid_values": ["eng", "spa"]} ] }

Size Thresholds Params

baseline_min_segment_size

Name	Description	Type	Default
baseline_min_segment_size	Minimal segment size for the baseline segment.	PositiveFloat	1

{ "baseline_min_segment_size": 100 }

benchmark_baseline_min_segment_size

Name	Description	Type	Default
benchmark_baseline_min_segment_size	Minimal segment size for the benchmark baseline segment.	PositiveFloat	1

{ "benchmark_baseline_min_segment_size": 100 }

benchmark_max_segment_size

Name	Description	Type	Default
benchmark_max_segment_size	Maximal benchmark segment size in number of records. Leave empty to not have such a threshold.	PositiveIntOrNone	None

{ "benchmark_max_segment_size": 1000 } benchmark_max_segment_size_fraction (NonInclusiveFractionOrNoneType) Maximal

benchmark_max_segment_size_fraction

Name	Description	Type	Default
benchmark_max_segment_size_fraction	Maximal benchmark segment size in fraction from baseline segment. Leave empty to not have such a threshold.	NonInclusiveFractionOrNone	None

{ "benchmark_max_segment_size_fraction": 0.2 }

benchmark_min_segment_size

Name	Description	Type	Default
benchmark_min_segment_size	Minimal benchmark segment size in number of records.	PositiveInt	100

{ "benchmark_min_segment_size": 50 }

benchmark_min_segment_size_fraction

Name	Description	Type	Default
benchmark_min_segment_size_fraction	Minimal benchmark segment size in fraction from baseline segment.	InclusiveFraction	0

{ "benchmark_min_segment_size_fraction": 0.05 }

max_segment_size

Name	Description	Type	Default
max_segment_size	Maximal segment size which a segment must have (bigger segments won't be considered in the search). Leave empty to not have such a threshold.	PositiveIntOrNone	None

{ "max_segment_size": 10000 }

max_segment_size_fraction

Name	Description	Type	Default
max_segment_size_fraction	Maximal segment size in fraction from baseline segment, which a segment must have. Leave empty to not have such a threshold.	NonInclusiveFractionOrNone	None

{ "max_segment_size_fraction": 0.2 }

min_segment_size

Name	Description	Type	Default
min_segment_size	Minimal segment size for the united benchmark+target segments.	PositiveInt	100

{ "min_segment_size": 100 }

min_segment_size_fraction

Name	Description	Type	Default
min_segment_size_fraction	Minimal segment size in fraction from baseline segment, which a segment must have in order to be considered in the search.	InclusiveFraction	0

{ "min_segment_size_fraction": 0.05 }

min_target_benchmark_size_ratio

Name	Description	Type	Default
min_target_benchmark_size_ratio	The minimum required ratio of segment size (after time range normalization) between target and benchmark periods.	InclusiveFraction	0.01

{ "min_target_benchmark_size_ratio": 0.05 }

target_baseline_min_segment_size

Name	Description	Type	Default
target_baseline_min_segment_size	Minimal segment size for the target baseline segment.	PositiveFloat	1

{ "target_baseline_min_segment_size": 0.8 }

target_max_segment_size

Name	Description	Type	Default
target_max_segment_size	Maximal target segment size in number of records. Leave empty to not have such a threshold.	PositiveIntOrNone	None

{ "target_max_segment_size": 10000 } target_max_segment_size_fraction (NonInclusiveFractionOrNoneType) Maximal target

target_max_segment_size_fraction

Name	Description	Type	Default
target_max_segment_size_fraction	Maximal target segment size in fraction from baseline segment. Leave empty to not have such a threshold.	NonInclusiveFractionOrNone	None

{ "target_max_segment_size_fraction": 0.2 }

target_min_segment_size

Name	Description	Type	Default
target_min_segment_size	Minimal target segment size in number of records.	PositiveInt	100

{ "target_min_segment_size": 50 } target_min_segment_size_fraction (InclusiveFractionType) Minimal target segment

target_min_segment_size_fraction

Name	Description	Type	Default
target_min_segment_size_fraction	Minimal target segment size in fraction from baseline segment.	InclusiveFraction	0

{ "target_min_segment_size_fraction": 0.05 }

Time Related Params

benchmark_set_period

Name	Description	Type	Default
benchmark_set_period	Time period for benchmark set. By default means the period just before the target set period. Expected format is "" where can be any positive integer, and options currently include: "d" (days), or "w" (weeks). e.g, "1d" means 1 day period.	TimePeriod

{ "benchmark_set_period": "50d" }

cadence

Name	Description	Type	Default
cadence	The cadence for evaluation of this verse. Only the following cadences are valid: Minutes: 1m, 5m, 10m, 15m, 20m, 30m. Hours: 1h, 2h, 3h, 4h, 6h, 8h, 12h. Days: 1d, 2d, 3d, 4d, 5d, 6d. Weeks: 1w, 2w, 3w, 4w, 5w.	CadenceType	1d

{ "cadence": "6h" }

expire_after

Name	Description	Type	Default
expire_after	Insights detected by this verse will continue to be considered active for at least this amount of time after the last time they were detected.	TimePeriod	3d

{ "expire_after": "2d" }

time_resolution

Name	Description	Type	Default
time_resolution	The time resolution of time series charts for insights of this verse. NOTE: this parameter doesn't affect the behavior of the insights finding mechanism, but only the data representation in the output.	VisualTimeResolution	1d

{ "time_resolution": "1w" }

Visuals and Enrichments Params

field_vectors

Name	Description	Type	Default
field_vectors	This attribute lists metric vectors for the FE to show on an insight card of this verse. A value in this field can either be a string (in which case the string should correspond to a kapi_vector name in the config) or an array (in which case the array should be treated as an ad-hoc kapi vector defined specifically for this verse).	FieldVectorsList	[]

{ "field_vectors": [ "field_vector_group_1", "field_vector_group_2", "field_vector_group_3" ] }

investigate_no_drill

Name	Description	Type	Default
investigate_no_drill	Dictates the link to the investigations page to add to the found insights. If True, the link will point to investigations page with a drilldown to the segment that was found. If it's false the link will point to the investigations page without drilldown, but with the found segment selected so it can be compared with a benchmark of a higher level.	bool	False

{ "investigate_no_drill": true }

time_resolution

Name	Description	Type	Default
time_resolution	The time resolution of time series charts for insights of this verse. NOTE: this parameter doesn't affect the behavior of the insights finding mechanism, but only the data representation in the output.	VisualTimeResolution	1d

{ "time_resolution": "1w" }