Overview

Overview of the dataset, analysis and alerts.

Dataset

Analysis

Alerts

isActive
55 (12%)
6 (1%)
371 (85%)
eyeColor
6 (1%)
3 (0%)
396 (97%)
age
7 (0%)
23 (3%)
726 (96%)

Histograms

This section contains visualisations of individual histograms and heatmaps of them over time.

Histogram Inspector

Heatmap

The heatmap shows the frequency of each value over time. If a variable has a high number of distinct values(i.e. has a high cardinality), then the most frequent values are displayed and the remaining are grouped as 'Others'. The maximum number of values to should is configurable (default: 20).

Column-Normalized Heatmap

The column-normalized heatmap allows for comparing of time bins when the counts in each bin vary.

Row-Normalized Heatmap

The row-normalized heatmaps allows for monitoring one value over time.

Histogram Inspector

Heatmap

The heatmap shows the frequency of each value over time. If a variable has a high number of distinct values(i.e. has a high cardinality), then the most frequent values are displayed and the remaining are grouped as 'Others'. The maximum number of values to should is configurable (default: 20).

Column-Normalized Heatmap

The column-normalized heatmap allows for comparing of time bins when the counts in each bin vary.

Row-Normalized Heatmap

The row-normalized heatmaps allows for monitoring one value over time.

Histogram Inspector

Traffic Lights

Traffic light calculation for different statistics (based on the calculated normalized residual, a.k.a. pull). Statistics for which all traffic lights are green are hidden from view by default.

Overview

count pull                           
distinct pull                           
entropy pull                           
filled pull                           
max pull                           
mean pull                           
mean trend10 zscore                           
min pull                           
most probable value pull                           
nan pull                           
overflow pull                           
p01 pull                           
p05 pull                           
p16 pull                           
p50 pull                           
p84 pull                           
p95 pull                           
p99 pull                           
prev1 chi2 zscore                           
prev1 ks zscore                           
ref chi2 zscore                           
ref jsd pull                           
ref ks zscore                           
ref max prob diff pull                           
ref psi pull                           
ref unknown labels                           
std pull                           
underflow pull                           
 2015-01-082015-01-222015-02-052015-02-192015-03-052015-03-192015-04-022015-04-162015-04-302015-05-142015-05-282015-06-112015-06-252015-07-092015-07-232015-08-062015-08-202015-09-032015-09-172015-10-012015-10-152015-10-292015-11-122015-11-262015-12-102015-12-242016-01-07

Overview

count pull                           
distinct pull                           
entropy pull                           
filled pull                           
nan pull                           
overflow pull                           
prev1 chi2 zscore                           
prev1 ks zscore                           
ref chi2 zscore                           
ref jsd pull                           
ref ks zscore                           
ref max prob diff pull                           
ref psi pull                           
ref unknown labels                           
underflow pull                           
 2015-01-082015-01-222015-02-052015-02-192015-03-052015-03-192015-04-022015-04-162015-04-302015-05-142015-05-282015-06-112015-06-252015-07-092015-07-232015-08-062015-08-202015-09-032015-09-172015-10-012015-10-152015-10-292015-11-122015-11-262015-12-102015-12-242016-01-07

Overview

count pull                           
distinct pull                           
entropy pull                           
filled pull                           
fraction of true pull                           
nan pull                           
overflow pull                           
prev1 chi2 zscore                           
prev1 ks zscore                           
ref chi2 zscore                           
ref jsd pull                           
ref ks zscore                           
ref max prob diff pull                           
ref psi pull                           
ref unknown labels                           
underflow pull                           
 2015-01-082015-01-222015-02-052015-02-192015-03-052015-03-192015-04-022015-04-162015-04-302015-05-142015-05-282015-06-112015-06-252015-07-092015-07-232015-08-062015-08-202015-09-032015-09-172015-10-012015-10-152015-10-292015-11-122015-11-262015-12-102015-12-242016-01-07

Alerts

Alerts aggregated by all traffic lights for each feature.

Overview

# green575757575756545557575756575755555457565657575754554742
# yellow000001320001000230110003177
# red2222222222222242222222223510
 2015-01-082015-01-222015-02-052015-02-192015-03-052015-03-192015-04-022015-04-162015-04-302015-05-142015-05-282015-06-112015-06-252015-07-092015-07-232015-08-062015-08-202015-09-032015-09-172015-10-012015-10-152015-10-292015-11-122015-11-262015-12-102015-12-242016-01-07

Overview

# green282828282828252628282827282828262628282728282825282218
# yellow000000320001000220010003045
# red000000000000000000000000025
 2015-01-082015-01-222015-02-052015-02-192015-03-052015-03-192015-04-022015-04-162015-04-302015-05-142015-05-282015-06-112015-06-252015-07-092015-07-232015-08-062015-08-202015-09-032015-09-172015-10-012015-10-152015-10-292015-11-122015-11-262015-12-102015-12-242016-01-07

Overview

# green151515151514151515151515151514151515151515151515141311
# yellow000001000000000000000000011
# red000000000000001000000000113
 2015-01-082015-01-222015-02-052015-02-192015-03-052015-03-192015-04-022015-04-162015-04-302015-05-142015-05-282015-06-112015-06-252015-07-092015-07-232015-08-062015-08-202015-09-032015-09-172015-10-012015-10-152015-10-292015-11-122015-11-262015-12-102015-12-242016-01-07

Overview

# green141414141414141414141414141413141314131414141414131213
# yellow000000000000000010100000121
# red222222222222223222222222222
 2015-01-082015-01-222015-02-052015-02-192015-03-052015-03-192015-04-022015-04-162015-04-302015-05-142015-05-282015-06-112015-06-252015-07-092015-07-232015-08-062015-08-202015-09-032015-09-172015-10-012015-10-152015-10-292015-11-122015-11-262015-12-102015-12-242016-01-07

Comparisons

Statistical comparisons of each time period (one bin) to the reference data.

Previous Reference

Comparing each time slot to the preceding time slot.

prev1 max prob diff

The largest absolute difference between all bin pairs of two normalized histograms

prev1 psi

Population Stability Index

prev1 jsd

Jensen-Shannon Divergence

prev1 ks

Kolmogorov-Smirnov test statistic

prev1 ks pvalue

p-value of the Kolmogorov-Smirnov test

prev1 ks zscore

Z-score of the Kolmogorov-Smirnov test

prev1 chi2

Chi-squared test statistic

prev1 chi2 norm

Normalized chi-squared statistic

prev1 chi2 zscore

Z-score of the chi-squared statistic

prev1 chi2 pvalue

p-value of the chi-squared statistic

prev1 chi2 max residual

The largest absolute normalized residual (|chi|) observed in all bin pairs

prev1 chi2 spike count

The number of normalized residuals of all bin pairs with absolute value bigger than a given threshold (default: 7).

External Reference

Comparing each time slot to the reference data.

ref max prob diff

The largest absolute difference between all bin pairs of two normalized histograms

ref psi

Population Stability Index

ref jsd

Jensen-Shannon Divergence

ref ks

Kolmogorov-Smirnov test statistic

ref ks pvalue

p-value of the Kolmogorov-Smirnov test

ref ks zscore

Z-score of the Kolmogorov-Smirnov test

ref chi2

Chi-squared test statistic

ref chi2 norm

Normalized chi-squared statistic

ref chi2 zscore

Z-score of the chi-squared statistic

ref chi2 pvalue

p-value of the chi-squared statistic

ref chi2 max residual

The largest absolute normalized residual (|chi|) observed in all bin pairs

ref chi2 spike count

The number of normalized residuals of all bin pairs with absolute value bigger than a given threshold (default: 7).

Others

mean trend10 zscore

Significance of (rolling) trend in means of features

Previous Reference

Comparing each time slot to the preceding time slot.

prev1 max prob diff

The largest absolute difference between all bin pairs of two normalized histograms

prev1 psi

Population Stability Index

prev1 jsd

Jensen-Shannon Divergence

prev1 unknown labels

Are categories observed in a given time slot that are not present in the reference?

prev1 chi2

Chi-squared test statistic

prev1 chi2 norm

Normalized chi-squared statistic

prev1 chi2 zscore

Z-score of the chi-squared statistic

prev1 chi2 pvalue

p-value of the chi-squared statistic

prev1 chi2 max residual

The largest absolute normalized residual (|chi|) observed in all bin pairs

prev1 chi2 spike count

The number of normalized residuals of all bin pairs with absolute value bigger than a given threshold (default: 7).

External Reference

Comparing each time slot to the reference data.

ref max prob diff

The largest absolute difference between all bin pairs of two normalized histograms

ref psi

Population Stability Index

ref jsd

Jensen-Shannon Divergence

ref unknown labels

Are categories observed in a given time slot that are not present in the reference?

ref chi2

Chi-squared test statistic

ref chi2 norm

Normalized chi-squared statistic

ref chi2 zscore

Z-score of the chi-squared statistic

ref chi2 pvalue

p-value of the chi-squared statistic

ref chi2 max residual

The largest absolute normalized residual (|chi|) observed in all bin pairs

ref chi2 spike count

The number of normalized residuals of all bin pairs with absolute value bigger than a given threshold (default: 7).

Previous Reference

Comparing each time slot to the preceding time slot.

prev1 max prob diff

The largest absolute difference between all bin pairs of two normalized histograms

prev1 psi

Population Stability Index

prev1 jsd

Jensen-Shannon Divergence

prev1 unknown labels

Are categories observed in a given time slot that are not present in the reference?

prev1 chi2

Chi-squared test statistic

prev1 chi2 norm

Normalized chi-squared statistic

prev1 chi2 pvalue

p-value of the chi-squared statistic

prev1 chi2 max residual

The largest absolute normalized residual (|chi|) observed in all bin pairs

prev1 chi2 spike count

The number of normalized residuals of all bin pairs with absolute value bigger than a given threshold (default: 7).

External Reference

Comparing each time slot to the reference data.

ref max prob diff

The largest absolute difference between all bin pairs of two normalized histograms

ref psi

Population Stability Index

ref jsd

Jensen-Shannon Divergence

ref unknown labels

Are categories observed in a given time slot that are not present in the reference?

ref chi2

Chi-squared test statistic

ref chi2 norm

Normalized chi-squared statistic

ref chi2 pvalue

p-value of the chi-squared statistic

ref chi2 max residual

The largest absolute normalized residual (|chi|) observed in all bin pairs

ref chi2 spike count

The number of normalized residuals of all bin pairs with absolute value bigger than a given threshold (default: 7).

Profiles

Basic statistics of the data (profiles) calculated for each time period (a period is represented by one bin). The yellow and red lines represent the corresponding traffic light bounds (default: 4 and 7 standard deviations with respect to the reference data).

min

Minimum value

max

Maximum value

p01

1% percentile

p05

5% percentile

p16

16% percentile

p50

50% percentile (median)

p84

84% percentile

p95

95% percentile

p99

99% percentile

mean

Mean value

std

Standard deviation

filled

Number of non-missing entries (non-NaN)

distinct

Number of distinct entries

most probable value

Most probable value

nan

Number of missing entries (NaN)

overflow

Number of values larger than the maximum bin-edge of the histogram.

underflow

Number of values smaller than the minimum bin-edge of the histogram.

count

Number of entries (non-NaN and NaN)

entropy

Entropy in nats

filled

Number of non-missing entries (non-NaN)

distinct

Number of distinct entries

most probable value

Most probable value

nan

Number of missing entries (NaN)

overflow

Number of values larger than the maximum bin-edge of the histogram.

underflow

Number of values smaller than the minimum bin-edge of the histogram.

count

Number of entries (non-NaN and NaN)

entropy

Entropy in nats

fraction of true

Compute fraction of 'true' (as in boolean) labels

filled

Number of non-missing entries (non-NaN)

distinct

Number of distinct entries

most probable value

Most probable value

nan

Number of missing entries (NaN)

overflow

Number of values larger than the maximum bin-edge of the histogram.

underflow

Number of values smaller than the minimum bin-edge of the histogram.

count

Number of entries (non-NaN and NaN)

entropy

Entropy in nats