About Wanalyzer Project / Statistics
WISDOMERA / WANALYZER

Wanalyzer Project / Statistics

Configuration of descriptive, comparative, before-and-after, and reliability analyses.

Analysis flow

Analysis type determines the relevant column lists and test design. Comparison-test items carry their own sample-size and distribution prerequisites, and results are interpreted with the selected significance threshold.

Dependency: Columns are mapped by index, so every analysis mapping must be checked after column order changes. Email fields are stored, but no delivery implementation was found in this backend scope.
status
status select

Activation value of the statistics definition. Record and list services carry the field, while automatic-analysis selection separately checks parameter 40. The inspected code contains no shared backend check proving that passive blocks every manual execution endpoint, so status alone must not be treated as a guarantee that execution is prevented.

Options and usage purposes
0 — passive
Stores the statistics definition as passive.
1 — active
Stores the statistics definition as active.
Example: 1 — aktif / active
Example:
statistics_config['status'] = '1'
Parameter ID: 0
wanalyzer project id
wanalyzer project id text

Client identifier connecting the statistics definition to its owning Wanalyzer project. With the correct identifier, the definition appears in the project statistics list and automatic-analysis scan; an incorrect identifier creates a relation under another project context.

Example: wp_01HXYZ…
Example:
statistics_config['wanalyzer_project_id'] = ''
Parameter ID: 1
name
name text

Name displayed for the analysis definition in lists and result screens. A descriptive and unique name distinguishes different test designs that use the same columns.

Example: Yaşa göre tedavi sonucu karşılaştırması
Example:
statistics_config['name'] = ''
Parameter ID: 2
note
note textarea

Note storing the purpose, population, and interpretation assumptions of the analysis. Code saves and returns it; a sufficient note clarifies which question the results answer.

Example: Birincil sonlanım için önceden tanımlanmış analiz.
Example:
statistics_config['note'] = ''
Parameter ID: 3
analyze type
analyze type select

Determines which statistical design group is used. Changing it makes the frontend display related fields such as descriptive columns, comparison tests, or before-and-after mappings; the design remains incomplete when lists required by the selected type are empty.

Options and usage purposes
0 — descriptive
Produces descriptive results for selected columns.
1 — cross columns
Designs comparisons with independent and dependent variables.
2 — before after comparison
Designs a before-and-after analysis between paired columns.
3 — cronbach alpha
Designs Cronbach alpha analysis for selected variables.
Example: 1 — karşılaştırma / cross columns
Example:
statistics_config['analyze_type'] = '1'
Parameter ID: 4
descriptive column selection type
descriptive column selection type select

Determines whether descriptive-analysis columns come from a manual list or are selected collectively by column type. Manual selection uses parameter 6; collective modes rely on column metadata.

Options and usage purposes
0 — manuel column selection
Uses only columns added to the list.
1 — all columns
Selects all eligible columns.
2 — all numeric columns
Selects numeric columns.
3 — all categorical columns
Selects categorical columns.
Example: 0 — manuel kolon seçimi / manuel column selection
Example:
statistics_config['descriptive_column_selection_type'] = '0'
Parameter ID: 5
descriptive selected column
descriptive selected column list_parameter

List containing column mappings used in descriptive analysis. Each item carries parameter 7, the column index, together with parameter 8, the usage type; indexes must be verified when column order changes.

Example: Kolon 2 — numeric
Example:
statistics_config['descriptive_selected_column'] = [{'descriptive_selected_column_index': '', 'descriptive_selected_column_usage_type': '0'}]
Parameter ID: 6
descriptive selected column index
descriptive selected column index text

Zero-based index of the project column used in the descriptive role. When the index matches the current column list, the intended data is analyzed; after columns are reordered, the mapping must be updated to avoid selecting a different column.

Example: 2
Example:
statistics_config['descriptive_selected_column'][0]['descriptive_selected_column_index'] = ''
Parameter ID: 7
descriptive selected column usage type
descriptive selected column usage type select

Determines whether the related column type is inferred automatically or forced to categorical or numeric. When forced, test selection and data preparation follow that role; a choice incompatible with the actual data may cause analysis failure or meaningless results.

Options and usage purposes
0 — auto
Requests automatic evaluation of the column type.
1 — categorical
Uses the column as a categorical variable.
2 — numeric
Uses the column as a numeric variable.
Example: 2 — numerik / numeric
Example:
statistics_config['descriptive_selected_column'][0]['descriptive_selected_column_usage_type'] = '2'
Parameter ID: 8
significance p value
significance p value float

P-value threshold used for statistical significance; the frontend defines 0.05 as the default. Because result p values are compared with this threshold, a smaller value produces a stricter and a larger value a more permissive significance decision.

Example: 0.05
Example:
statistics_config['significance_p_value'] = ''
Parameter ID: 9
comparison multi dimension
comparison multi dimension select

Stores whether the comparison is designed as multidimensional. Yes declares a design that considers multiple variable dimensions together; actual applicability must be evaluated with the selected test and column lists.

Options and usage purposes
0 — no
Disables the related behavior.
1 — yes
Enables the related behavior.
Example: 0 — hayır / no
Example:
statistics_config['comparison_multi_dimension'] = '0'
Parameter ID: 10
independent variable column
independent variable column list_parameter

List containing independent variable columns. Each item stores a column index and usage type; list order and indexes directly determine analysis input.

Example: age, gender
Example:
statistics_config['independent_variable_column'] = [{'independent_variable_column_index': '', 'independent_variable_column_usage_type': '0'}]
Parameter ID: 11
independent variable column index
independent variable column index text

Zero-based index of the project column used in the independent-variable role. When the index matches the current column list, the intended data is analyzed; after columns are reordered, the mapping must be updated to avoid selecting a different column.

Example: 2
Example:
statistics_config['independent_variable_column'][0]['independent_variable_column_index'] = ''
Parameter ID: 12
independent variable column usage type
independent variable column usage type select

Determines whether the related column type is inferred automatically or forced to categorical or numeric. When forced, test selection and data preparation follow that role; a choice incompatible with the actual data may cause analysis failure or meaningless results.

Options and usage purposes
0 — auto
Requests automatic evaluation of the column type.
1 — categorical
Uses the column as a categorical variable.
2 — numeric
Uses the column as a numeric variable.
Example: 2 — numerik / numeric
Example:
statistics_config['independent_variable_column'][0]['independent_variable_column_usage_type'] = '2'
Parameter ID: 13
dependent variable column
dependent variable column list_parameter

List containing dependent variable columns. Each item stores a column index and usage type, and may additionally contain a categorical target category; list order and indexes directly determine analysis input.

Example: outcome
Example:
statistics_config['dependent_variable_column'] = [{'dependent_variable_column_index': '', 'dependent_variable_column_usage_type': '0', 'dependent_variable_column_option_target_category': ''}]
Parameter ID: 14
dependent variable column index
dependent variable column index text

Zero-based index of the project column used in the dependent-variable role. When the index matches the current column list, the intended data is analyzed; after columns are reordered, the mapping must be updated to avoid selecting a different column.

Example: 7
Example:
statistics_config['dependent_variable_column'][0]['dependent_variable_column_index'] = ''
Parameter ID: 15
dependent variable column usage type
dependent variable column usage type select

Determines whether the related column type is inferred automatically or forced to categorical or numeric. When forced, test selection and data preparation follow that role; a choice incompatible with the actual data may cause analysis failure or meaningless results.

Options and usage purposes
0 — auto
Requests automatic evaluation of the column type.
1 — categorical
Uses the column as a categorical variable.
2 — numeric
Uses the column as a numeric variable.
Example: 1 — kategorik / categorical
Example:
statistics_config['dependent_variable_column'][0]['dependent_variable_column_usage_type'] = '1'
Parameter ID: 16
dependent variable column option target category
dependent variable column option target category text

Option value treated as the success or event category for a categorical dependent variable. Particularly in logistic regression, selecting the wrong target category may make the model explain the opposite event.

Example: success
Example:
statistics_config['dependent_variable_column'][0]['dependent_variable_column_option_target_category'] = ''
Parameter ID: 17
auto comparison variable column
auto comparison variable column list_parameter

List containing automatic-comparison variable columns. Each item stores a column index and usage type; list order and indexes directly determine analysis input.

Example: age, score
Example:
statistics_config['auto_comparison_variable_column'] = [{'auto_comparison_variable_column_index': '', 'auto_comparison_variable_column_usage_type': '0'}]
Parameter ID: 18
auto comparison variable column index
auto comparison variable column index text

Zero-based index of the project column used in the automatic-comparison role. When the index matches the current column list, the intended data is analyzed; after columns are reordered, the mapping must be updated to avoid selecting a different column.

Example: 4
Example:
statistics_config['auto_comparison_variable_column'][0]['auto_comparison_variable_column_index'] = ''
Parameter ID: 19
auto comparison variable column usage type
auto comparison variable column usage type select

Determines whether the related column type is inferred automatically or forced to categorical or numeric. When forced, test selection and data preparation follow that role; a choice incompatible with the actual data may cause analysis failure or meaningless results.

Options and usage purposes
0 — auto
Requests automatic evaluation of the column type.
1 — categorical
Uses the column as a categorical variable.
2 — numeric
Uses the column as a numeric variable.
Example: 0 — otomatik / auto
Example:
statistics_config['auto_comparison_variable_column'][0]['auto_comparison_variable_column_usage_type'] = '0'
Parameter ID: 20
comparison test
comparison test list_parameter

List containing comparison-test definitions. Each item combines a test name with pre-analysis sample-size, distribution, and independent-variable usage checks, allowing different tests to run under different eligibility rules.

Example: Student t-test; n ≥ 30
Example:
statistics_config['comparison_test'] = [{'comparison_test_name': '', 'before_analyze_check_type': '0', 'total_sample_count_check_type': '', 'total_sample_count': '', 'group_sample_count_check_type': '', 'group_sample_count': '', 'data_distribution_test_usage_type': '', 'data_distribution_skewness_usage': '', 'data_distribution_skewness_value': '', 'data_distribution_kurtosis_usage': '', 'data_distribution_kurtosis_value': '', 'data_distribution_shapiro_wilk_usage': '', 'data_distribution_kolmogorov_smirnov_usage': '', 'independent_variable_usage_type': ''}]
Parameter ID: 21
comparison test name
comparison test name select

Selects the statistical test in a comparison-list item. Test assumptions must match categorical or numeric column roles; an incompatible test may fail or produce an uninterpretable result.

Options and usage purposes
0 — oneway anova
Adds this test to the comparison definition.
1 — chisquare
Adds this test to the comparison definition.
2 — fisher exact
Adds this test to the comparison definition.
3 — pearson correlation
Adds this test to the comparison definition.
4 — spearman correlation
Adds this test to the comparison definition.
5 — logistic regression
Adds this test to the comparison definition.
6 — linear regression
Adds this test to the comparison definition.
7 — kruskal wallis
Adds this test to the comparison definition.
8 — mann whitney u
Adds this test to the comparison definition.
9 — independent samples student t-test
Adds this test to the comparison definition.
10 — welch t test
Adds this test to the comparison definition.
Example: 9 — independent samples student t-test
Example:
statistics_config['comparison_test'][0]['comparison_test_name'] = '9'
Parameter ID: 22
before analyze check type
before analyze check type select

Determines which sample-size and distribution checks, and which AND/OR combination, are required before a test runs. The selection reveals dependent threshold and distribution fields; when the condition fails, the test is intended not to be used.

Options and usage purposes
0 — not control
Applies this validation combination before the test.
1 — only total sample size
Applies this validation combination before the test.
2 — only every group sample size
Applies this validation combination before the test.
3 — only total samples data distribution
Applies this validation combination before the test.
4 — only every group data distribution
Applies this validation combination before the test.
5 — both total sample size and data distribution
Applies this validation combination before the test.
6 — both every group sample size and data distribution
Applies this validation combination before the test.
7 — total sample size or data distribution
Applies this validation combination before the test.
8 — every group sample size or data distribution
Applies this validation combination before the test.
Example: 5
Example:
statistics_config['comparison_test'][0]['before_analyze_check_type'] = '5'
Parameter ID: 23
total sample count check type
total sample count check type select

Determines which comparison operator tests the actual sample count against the configured threshold. Reversing the operator may reject sufficient data or accept insufficient data.

Options and usage purposes
0 — smaller
Tests whether the actual count is smaller than the threshold.
1 — smaller and equal
Tests whether the actual count is smaller than or equal to the threshold.
2 — bigger
Tests whether the actual count is greater than the threshold.
3 — bigger and equal
Tests whether the actual count is greater than or equal to the threshold.
Example: 3 — büyük eşittir / bigger and equal
Example:
statistics_config['comparison_test'][0]['total_sample_count_check_type'] = '3'
Parameter ID: 24
total sample count
total sample count integer

Sample-count threshold required for the total. It is meaningful only when the corresponding pre-analysis check type is selected; use a positive value appropriate to the study design.

Example: 30
Example:
statistics_config['comparison_test'][0]['total_sample_count'] = ''
Parameter ID: 25
group sample count check type
group sample count check type select

Determines which comparison operator tests the actual sample count against the configured threshold. Reversing the operator may reject sufficient data or accept insufficient data.

Options and usage purposes
0 — smaller
Tests whether the actual count is smaller than the threshold.
1 — smaller and equal
Tests whether the actual count is smaller than or equal to the threshold.
2 — bigger
Tests whether the actual count is greater than the threshold.
3 — bigger and equal
Tests whether the actual count is greater than or equal to the threshold.
Example: 3 — büyük eşittir / bigger and equal
Example:
statistics_config['comparison_test'][0]['group_sample_count_check_type'] = '3'
Parameter ID: 26
group sample count
group sample count integer

Sample-count threshold required for each group. It is meaningful only when the corresponding pre-analysis check type is selected; use a positive value appropriate to the study design.

Example: 15
Example:
statistics_config['comparison_test'][0]['group_sample_count'] = ''
Parameter ID: 27
data distribution test usage type
data distribution test usage type select

Determines how results of multiple distribution checks are combined. One, all, or all-not-normal choices produce different eligibility decisions from the same test results.

Options and usage purposes
0 — one of them
Passes when one selected check supports normality.
1 — all of them
Requires every selected check to support normality.
2 — all of them not normally distributed
Requires every selected check to reject normality.
Example: 1 — all of them
Example:
statistics_config['comparison_test'][0]['data_distribution_test_usage_type'] = '1'
Parameter ID: 28
data distribution skewness usage
data distribution skewness usage select

Determines whether the related distribution measure or test participates in pre-analysis evaluation. With yes, its result is included under the combination rule in parameter 28; with no, it does not contribute to the decision.

Options and usage purposes
0 — no
Disables the related behavior.
1 — yes
Enables the related behavior.
Example: 1 — evet / yes
Example:
statistics_config['comparison_test'][0]['data_distribution_skewness_usage'] = '1'
Parameter ID: 29
data distribution skewness value
data distribution skewness value float

Acceptance threshold for skewness. Unless the related usage field is yes, the value does not participate in the decision; widening the threshold accepts more distributions.

Example: 2.0
Example:
statistics_config['comparison_test'][0]['data_distribution_skewness_value'] = ''
Parameter ID: 30
data distribution kurtosis usage
data distribution kurtosis usage select

Determines whether the related distribution measure or test participates in pre-analysis evaluation. With yes, its result is included under the combination rule in parameter 28; with no, it does not contribute to the decision.

Options and usage purposes
0 — no
Disables the related behavior.
1 — yes
Enables the related behavior.
Example: 1 — evet / yes
Example:
statistics_config['comparison_test'][0]['data_distribution_kurtosis_usage'] = '1'
Parameter ID: 31
data distribution kurtosis value
data distribution kurtosis value float

Acceptance threshold for kurtosis. Unless the related usage field is yes, the value does not participate in the decision; widening the threshold accepts more distributions.

Example: 7.0
Example:
statistics_config['comparison_test'][0]['data_distribution_kurtosis_value'] = ''
Parameter ID: 32
data distribution shapiro wilk usage
data distribution shapiro wilk usage select

Determines whether the related distribution measure or test participates in pre-analysis evaluation. With yes, its result is included under the combination rule in parameter 28; with no, it does not contribute to the decision.

Options and usage purposes
0 — no
Disables the related behavior.
1 — yes
Enables the related behavior.
Example: 1 — evet / yes
Example:
statistics_config['comparison_test'][0]['data_distribution_shapiro_wilk_usage'] = '1'
Parameter ID: 33
data distribution kolmogorov smirnov usage
data distribution kolmogorov smirnov usage select

Determines whether the related distribution measure or test participates in pre-analysis evaluation. With yes, its result is included under the combination rule in parameter 28; with no, it does not contribute to the decision.

Options and usage purposes
0 — no
Disables the related behavior.
1 — yes
Enables the related behavior.
Example: 0 — hayır / no
Example:
statistics_config['comparison_test'][0]['data_distribution_kolmogorov_smirnov_usage'] = '0'
Parameter ID: 34
independent variable usage type
independent variable usage type select

Determines whether multiple independent variables are evaluated separately, jointly, or in two stages. The two-stage choice carries variables significant in individual analysis into a joint model.

Options and usage purposes
0 — one by one
Evaluates independent variables in separate models.
1 — all together
Evaluates independent variables in one model.
2 — first one by one then all the meaningful ones together
Runs individual evaluation first and carries significant variables into a joint model.
Example: 2
Example:
statistics_config['comparison_test'][0]['independent_variable_usage_type'] = '2'
Parameter ID: 35
before-after test
before-after test list_parameter

List containing before-and-after analysis pairs. Each item contains a test name and before and after column indexes belonging to the same observations; using unpaired columns violates the paired-test assumption.

Example: before_score → after_score
Example:
statistics_config['before_after_test'] = [{'before_after_test_name': '', 'before_after_1st_column_index': '', 'before_after_2nd_column_index': ''}]
Parameter ID: 36
before-after test name
before-after test name select

Selects the paired test applied to the before and after column pair. The current option is the paired-samples t test and requires numeric, matched observations.

Options and usage purposes
0 — paired samples t test
Selects the paired-samples t test for two paired columns.
Example: 0 — paired samples t test
Example:
statistics_config['before_after_test'][0]['before_after_test_name'] = '0'
Parameter ID: 37
before-after 1st column index
before-after 1st column index text

Zero-based index of the project column used in the before role. When the index matches the current column list, the intended data is analyzed; after columns are reordered, the mapping must be updated to avoid selecting a different column.

Example: 3
Example:
statistics_config['before_after_test'][0]['before_after_1st_column_index'] = ''
Parameter ID: 38
before-after 2nd column index
before-after 2nd column index text

Zero-based index of the project column used in the after role. When the index matches the current column list, the intended data is analyzed; after columns are reordered, the mapping must be updated to avoid selecting a different column.

Example: 4
Example:
statistics_config['before_after_test'][0]['before_after_2nd_column_index'] = ''
Parameter ID: 39
automatic analysis after project data preparation
automatic analysis after project data preparation select

Determines whether this statistics definition becomes eligible for automatic execution after project data preparation. AutomaticAnalyzeProcess includes the definition only when the value is yes; with no, manual execution may remain available but no post-preparation automatic trigger occurs.

Options and usage purposes
0 — no
Disables the related behavior.
1 — yes
Enables the related behavior.
Example: 1 — evet / yes
Example:
statistics_config['automatic_analysis_after_project_data_preparation'] = '1'
Parameter ID: 40
after analyze send mail
after analyze send mail select

Stores the request to send email after analysis, and the frontend displays mail fields 42 through 58 when it is yes. No implementation using these fields to send email was found in the inspected Wanalyzer backend scope, so delivery may depend on another service.

Options and usage purposes
0 — no
Disables the related behavior.
1 — yes
Enables the related behavior.
Example: 0 — hayır / no
Example:
statistics_config['after_analyze_send_mail'] = '0'
Parameter ID: 41
mail attachment file types
mail attachment file types checkbox

Lists file formats requested as analysis-email attachments. Excel and PDF may both be selected; record and get services carry the field, but attachment generation and delivery were not verified in the inspected backend scope.

Options and usage purposes
0 — excel
Selects an Excel attachment.
1 — pdf
Selects a PDF attachment.
Example: excel, pdf
Example:
statistics_config['mail_attachment_file_types'] = []
Parameter ID: 42
mail list
mail list textarea

Stores the analysis email’s recipient addresses. Statistics record and get services preserve the field; because its interpretation during email rendering was not verified in the inspected Wanalyzer backend scope, format, color, or dimension values must match the target mail service.

Example: analyst@example.com
Example:
statistics_config['mail_list'] = ''
Parameter ID: 43
mail sender name
mail sender name textarea

Stores the analysis email’s displayed sender name. Statistics record and get services preserve the field; because its interpretation during email rendering was not verified in the inspected Wanalyzer backend scope, format, color, or dimension values must match the target mail service.

Example: WisdomEra Analiz
Example:
statistics_config['mail_sender_name'] = ''
Parameter ID: 44
mail subject prefix
mail subject prefix textarea

Stores the analysis email’s subject prefix. Statistics record and get services preserve the field; because its interpretation during email rendering was not verified in the inspected Wanalyzer backend scope, format, color, or dimension values must match the target mail service.

Example: [Wanalyzer]
Example:
statistics_config['mail_subject_prefix'] = ''
Parameter ID: 45
mail subject suffix
mail subject suffix textarea

Stores the analysis email’s subject suffix. Statistics record and get services preserve the field; because its interpretation during email rendering was not verified in the inspected Wanalyzer backend scope, format, color, or dimension values must match the target mail service.

Example: — Sonuçlar
Example:
statistics_config['mail_subject_suffix'] = ''
Parameter ID: 46
mail start note
mail start note textarea

Stores the analysis email’s message introduction. Statistics record and get services preserve the field; because its interpretation during email rendering was not verified in the inspected Wanalyzer backend scope, format, color, or dimension values must match the target mail service.

Example: Analiz sonuçları aşağıdadır.
Example:
statistics_config['mail_start_note'] = ''
Parameter ID: 47
mail note
mail note list_parameter

Stores the analysis email’s repeatable content blocks. Statistics record and get services preserve the field; because its interpretation during email rendering was not verified in the inspected Wanalyzer backend scope, format, color, or dimension values must match the target mail service.

Example: Önemli sonuçlar
Example:
statistics_config['mail_note'] = [{'mail_caption': '', 'mail_content': '', 'mail_content_color': ''}]
Parameter ID: 48
mail caption
mail caption text

Stores the analysis email’s content caption. Statistics record and get services preserve the field; because its interpretation during email rendering was not verified in the inspected Wanalyzer backend scope, format, color, or dimension values must match the target mail service.

Example: Özet
Example:
statistics_config['mail_note'][0]['mail_caption'] = ''
Parameter ID: 49
mail content
mail content textarea

Stores the analysis email’s content text. Statistics record and get services preserve the field; because its interpretation during email rendering was not verified in the inspected Wanalyzer backend scope, format, color, or dimension values must match the target mail service.

Example: p < 0.05 olan sonuçlar gösterilmiştir.
Example:
statistics_config['mail_note'][0]['mail_content'] = ''
Parameter ID: 50
mail content color
mail content color text

Stores the analysis email’s content color. Statistics record and get services preserve the field; because its interpretation during email rendering was not verified in the inspected Wanalyzer backend scope, format, color, or dimension values must match the target mail service.

Example: #17344a
Example:
statistics_config['mail_note'][0]['mail_content_color'] = ''
Parameter ID: 51
mail background image
mail background image image

Stores the analysis email’s background image. Statistics record and get services preserve the field; because its interpretation during email rendering was not verified in the inspected Wanalyzer backend scope, format, color, or dimension values must match the target mail service.

Example: data:image/png;base64,…
Example:
statistics_config['mail_background_image'] = ''
Parameter ID: 52
mail table border color
mail table border color text

Stores the analysis email’s table-border color. Statistics record and get services preserve the field; because its interpretation during email rendering was not verified in the inspected Wanalyzer backend scope, format, color, or dimension values must match the target mail service.

Example: #dfe5f0
Example:
statistics_config['mail_table_border_color'] = ''
Parameter ID: 53
mail table background color
mail table background color textarea

Stores the analysis email’s table-background color. Statistics record and get services preserve the field; because its interpretation during email rendering was not verified in the inspected Wanalyzer backend scope, format, color, or dimension values must match the target mail service.

Example: #ffffff
Example:
statistics_config['mail_table_background_color'] = ''
Parameter ID: 54
mail signature
mail signature list_parameter

Stores the analysis email’s signature blocks. Statistics record and get services preserve the field; because its interpretation during email rendering was not verified in the inspected Wanalyzer backend scope, format, color, or dimension values must match the target mail service.

Example: Analiz Ekibi
Example:
statistics_config['mail_signature'] = [{'mail_signature_note': '', 'mail_signature_image': '', 'mail_signature_width': ''}]
Parameter ID: 55
mail signature note
mail signature note textarea

Stores the analysis email’s signature note. Statistics record and get services preserve the field; because its interpretation during email rendering was not verified in the inspected Wanalyzer backend scope, format, color, or dimension values must match the target mail service.

Example: Saygılarımızla
Example:
statistics_config['mail_signature'][0]['mail_signature_note'] = ''
Parameter ID: 56
mail signature image
mail signature image image

Stores the analysis email’s signature image. Statistics record and get services preserve the field; because its interpretation during email rendering was not verified in the inspected Wanalyzer backend scope, format, color, or dimension values must match the target mail service.

Example: data:image/png;base64,…
Example:
statistics_config['mail_signature'][0]['mail_signature_image'] = ''
Parameter ID: 57
mail signature width
mail signature width textarea

Stores the analysis email’s signature-image width. Statistics record and get services preserve the field; because its interpretation during email rendering was not verified in the inspected Wanalyzer backend scope, format, color, or dimension values must match the target mail service.

Example: 160px
Example:
statistics_config['mail_signature'][0]['mail_signature_width'] = ''
Parameter ID: 58
column filter rule json
column filter rule json textarea

JSON rule restricting dataset rows included in the analysis by column conditions. The frontend rule editor produces query_list and compilation arrays, the record service serializes the object to JSON text, and the get service parses it back; manually entered invalid JSON may prevent the definition from loading.

Example: Yaş ≥ 18
Example:
statistics_config['column_filter_rule_json'] = {'query_list': [], 'compile_str_array': [], 'compile_str_date': [], 'compile_str_text': []}
Parameter ID: 59
comparison auto combination analyze
comparison auto combination analyze select

Enables or disables automatic generation of variable combinations for comparison analysis. The frontend displays it only for comparison designs; active carries the request to generate combinations from column lists, while the number of combinations depends on selected columns.

Options and usage purposes
0 — passive
Disables automatic combination generation.
1 — active
Enables automatic combination generation.
Example: 1 — aktif / active
Example:
statistics_config['comparison_auto_combination_analyze'] = '1'
Parameter ID: 60
About Wanalyzer Project / Statistics / Process
WISDOMERA / WANALYZER

Wanalyzer Project / Statistics / Process

Tracks the status, times, counters, and log of every statistics execution.

Process flow

When analysis is triggered, the process is created with start status and zero counters. Start and continue are actively tracked; cancel and complete end tracking and refresh results.

Layer separation: Result status, test name, analysis type, and result-text language are not parameters of this process; they are stored in the process-result child layer.
status
status select

Lifecycle status of the statistical analysis process. The record is created with start when analysis is triggered. When start or continue is selected, the process is treated as active and the interface refreshes it every five seconds; when cancel or complete is selected, tracking stops and the result list is reloaded.

Options and usage purposes
0 — cancel
Removes the process from active tracking and makes the interface refresh results.
1 — start
A new process is created with this value and tracked as active.
2 — continue
Indicates that execution is in progress and continues active tracking.
3 — complete
Indicates that execution has finished, stops tracking, and refreshes results.
Example: 1 — başla / start
Example:
statistics_process_config['status'] = '1'
Parameter ID: 0
wanalyzer project id
wanalyzer project id text

Client identifier connecting the process to its owning Wanalyzer project. With the correct project identifier, the active-process service reports the record under the related project and result authorization is evaluated in the correct project context; an incorrect identifier breaks the process-to-project relationship.

Example: wp_01HXYZ…
Example:
statistics_process_config['wanalyzer_project_id'] = ''
Parameter ID: 1
stats id
stats id text

Client identifier of the statistics definition that produced the process. The process list is read under the selected statistics record through this relation; the correct identifier associates generated results with the correct analysis design, while an incorrect identifier may make them appear under another statistics definition.

Example: stats_01HXYZ…
Example:
statistics_process_config['stats_id'] = ''
Parameter ID: 2
start date
start date datetime

Start time at which the analysis process is created. The trigger service fills this field in Coordinated Universal Time with second precision; it is therefore used to order processes chronologically and to calculate approximate execution duration together with the end date.

Example: 2026-10-25T09:30:00
Example:
statistics_process_config['start_date'] = ''
Parameter ID: 3
end date
end date datetime

Time at which analysis execution ends. When populated after process completion, it defines the execution interval together with the start date; it may remain empty while the process is still running or when a cancelled record has not produced an end time.

Example: 2026-10-25T09:31:42
Example:
statistics_process_config['end_date'] = ''
Parameter ID: 4
last information date
last information date datetime

Shows when the running process most recently reported progress. It is initialized to the same value as the start date and the active-process service returns it together with status; regular updates indicate process activity, while a value that remains unchanged for a long time may indicate a delayed or stalled process.

Example: 2026-10-25T09:30:45
Example:
statistics_process_config['last_information_date'] = ''
Parameter ID: 5
stop date
stop date datetime

Stores the time at which the process was stopped or cancelled. When populated together with cancel status, it explains when execution was terminated; a value in this field alone does not stop interface tracking because the tracking decision depends on status being cancel or complete.

Example: 2026-10-25T09:30:58
Example:
statistics_process_config['stop_date'] = ''
Parameter ID: 6
row count
row count integer

Counter holding the number of dataset rows associated with or processed by the analysis process. It is initialized to zero when the process record is created; when the analysis engine updates it, it indicates scope and progress, while remaining zero does not by itself mean failure because the process may have only just started.

Example: 1250
Example:
statistics_process_config['row_count'] = ''
Parameter ID: 7
result total count
result total count integer

Counter holding the total number of result records produced by the process. It is initialized to zero when the record is created; when updated as results are written to the child result layer, it allows the user to see the expected result volume. This field is not the result content; details are read from the process-result layer.

Example: 12
Example:
statistics_process_config['result_total_count'] = ''
Parameter ID: 8
log
log textarea

Text carrying the process launch mode, progress notes, or error descriptions. At creation, the trigger service writes an Analyze entry identifying authorization or credit execution mode; when subsequent steps append details, this field can explain why the operation completed, was delayed, or was cancelled.

Example: Analyze (auth)
Example:
statistics_process_config['log'] = ''
Parameter ID: 9
About Wanalyzer Project / Statistics / Process / Result
WISDOMERA / WANALYZER

Wanalyzer Project / Statistics / Process / Result

Stores measures, column relations, diagnostics, and tables of analysis results.

Result flow

Analysis and test type determine field interpretation; the p value is compared with the significance threshold; raw JSON fields generate tables and distribution diagnostics.

Dependency: Column indexes depend on project order and JSON fields depend on the expected key structure. Changing either structure may break presentation of historical results.
status
status select

Activation status of the result record. Active marks the record as suitable for use as a current result, while passive marks it as historical or out of use. The inspected list service returns child records without applying a status filter, so this value alone must not be assumed to hide the record from lists.

Options and usage purposes
0 — passive
Marks the result record as passive.
1 — active
Marks the result record as active.
Example: 1 — aktif / active
Example:
statistics_result_config['status'] = '1'
Parameter ID: 0
wanalyzer project id
wanalyzer project id text

Client identifier connecting the result record to the related project. With the correct identifier, the result is found in the correct hierarchy; an incorrect identifier may direct authorization, listing, or historical-result selection to a different context.

Example: wp_01HXYZ…
Example:
statistics_result_config['wanalyzer_project_id'] = ''
Parameter ID: 1
stats id
stats id text

Client identifier connecting the result record to the related statistics definition. With the correct identifier, the result is found in the correct hierarchy; an incorrect identifier may direct authorization, listing, or historical-result selection to a different context.

Example: stats_01HXYZ…
Example:
statistics_result_config['stats_id'] = ''
Parameter ID: 2
process id
process id text

Client identifier connecting the result record to the related statistics process. With the correct identifier, the result is found in the correct hierarchy; an incorrect identifier may direct authorization, listing, or historical-result selection to a different context.

Example: process_01HXYZ…
Example:
statistics_result_config['process_id'] = ''
Parameter ID: 3
start date
start date datetime

Time when result generation started. Together, start and end dates allow execution duration to be calculated.

Example: 2026-10-25T09:30:00
Example:
statistics_result_config['start_date'] = ''
Parameter ID: 4
end date
end date datetime

Time when result generation finished. When evaluated with the start date, the end date gives execution duration; if empty, the result may be unfinished or the date may not have been recorded.

Example: 2026-10-25T09:30:04
Example:
statistics_result_config['end_date'] = ''
Parameter ID: 5
analyze type
analyze type select

Determines which analysis design produced the result. Because it controls how the interface interprets related fields such as descriptive columns, comparison variables, before-and-after mappings, or the Cronbach alpha measure, it must agree with the result content.

Options and usage purposes
0 — descriptive
Reports distribution and summary measures for one column.
1 — cross columns
Reports the test between independent and dependent columns.
2 — before after comparison
Reports paired before and after columns.
3 — cronbach alpha
Reports an internal-consistency analysis result.
Example: 1 — karşılaştırma / cross columns
Example:
statistics_result_config['analyze_type'] = '1'
Parameter ID: 6
test name
test name select

Identifies the statistical test that produced a comparative result. The interface prioritizes median and interquartile range for nonparametric Kruskal–Wallis and Mann–Whitney U results, and mean and standard deviation for other comparisons, so the correct test code directly affects presentation.

Options and usage purposes
0 — oneway anova
Indicates that the result was produced with this statistical test.
1 — chisquare
Indicates that the result was produced with this statistical test.
2 — fisher exact
Indicates that the result was produced with this statistical test.
3 — pearson correlation
Indicates that the result was produced with this statistical test.
4 — spearman correlation
Indicates that the result was produced with this statistical test.
5 — logistic regression
Indicates that the result was produced with this statistical test.
6 — linear regression
Indicates that the result was produced with this statistical test.
7 — kruskal wallis
Indicates that the result was produced with this statistical test.
8 — mann whitney u
Indicates that the result was produced with this statistical test.
9 — independent samples student t-test
Indicates that the result was produced with this statistical test.
10 — welch t test
Indicates that the result was produced with this statistical test.
Example: 9 — independent samples student t-test
Example:
statistics_result_config['test_name'] = '9'
Parameter ID: 7
row count
row count integer

Number of valid observations used to calculate the result. It may serve as the denominator for percentages and table totals; when empty or zero in older records, the interface uses the sum of counts in frequency JSON as a fallback denominator.

Example: 120
Example:
statistics_result_config['row_count'] = ''
Parameter ID: 8
p value
p value text

Calculated p value of the statistical test. The interface compares it with the significance threshold in parameter 10 and highlights the result as significant when the p value is smaller. Values below 0.001 are displayed as “<0.001”.

Example: 0.032
Example:
statistics_result_config['p_value'] = ''
Parameter ID: 9
significance p value
significance p value text

P-value threshold used to decide whether the result is significant. If the field cannot be converted to a number, test-selection assessment uses 0.05; a smaller threshold gives a stricter decision and a larger threshold gives a more permissive decision.

Example: 0.05
Example:
statistics_result_config['significance_p_value'] = ''
Parameter ID: 10
log
log textarea

Explanation, warning, or error note produced by the analysis engine. The interface carries this field as the result note in categorical-to-numeric group tables, so calculation limitations or reasons for missing data should be stated clearly here.

Example: Analiz 120 geçerli gözlemle tamamlandı.
Example:
statistics_result_config['log'] = ''
Parameter ID: 11
result text
result text list_parameter

List carrying interpretation text for the same result in multiple languages. Each item stores parameter 13 content together with parameter 14 language; when an item matching the interface language exists, the user sees a readable interpretation together with calculated values.

Example: Türkçe ve İngilizce yorum
Example:
statistics_result_config['result_text'] = [{'result_text_content': '', 'result_text_language': '0'}]
Parameter ID: 12
result text content
result text content textarea

Readable interpretation content in a result-text item. Explaining the p value or effect direction in context reduces the risk of misinterpreting the raw number; content that does not match its language field may be presented under the wrong language.

Example: Gruplar arasında anlamlı fark bulundu.
Example:
statistics_result_config['result_text'][0]['result_text_content'] = ''
Parameter ID: 13
result text language
result text language select

Language code of a result-text item. Selecting Turkish or English determines the language view in which the content is used; an incorrect code may prevent the interface from selecting the correct text.

Options and usage purposes
0 — tr
Indicates that the text is the Turkish version.
1 — en
Indicates that the text is the English version.
Example: 0 — tr
Example:
statistics_result_config['result_text'][0]['result_text_language'] = '0'
Parameter ID: 14
descriptive selected column index
descriptive selected column index text

Zero-based index of the project column used in the descriptive role. Because the interface resolves the column name and options from this index, reordering project columns later may associate an old result with a different column or make the table skip it.

Example: 2
Example:
statistics_result_config['descriptive_selected_column_index'] = ''
Parameter ID: 15
independent variable column
independent variable column list_parameter

List carrying independent variable columns used in the result. Each item contains its column index; because axis and group names for categorical-to-categorical and categorical-to-numeric tables are produced from these mappings, the list must represent the same roles as the result JSON structure.

Example: age
Example:
statistics_result_config['independent_variable_column'] = [{'independent_variable_column_index': ''}]
Parameter ID: 16
independent variable column index
independent variable column index text

Zero-based index of the project column used in the independent-variable role. Because the interface resolves the column name and options from this index, reordering project columns later may associate an old result with a different column or make the table skip it.

Example: 2
Example:
statistics_result_config['independent_variable_column'][0]['independent_variable_column_index'] = ''
Parameter ID: 17
dependent variable column
dependent variable column list_parameter

List carrying dependent variable columns used in the result. Each item contains its column index; because axis and group names for categorical-to-categorical and categorical-to-numeric tables are produced from these mappings, the list must represent the same roles as the result JSON structure.

Example: outcome
Example:
statistics_result_config['dependent_variable_column'] = [{'dependent_variable_column_index': ''}]
Parameter ID: 18
dependent variable column index
dependent variable column index text

Zero-based index of the project column used in the dependent-variable role. Because the interface resolves the column name and options from this index, reordering project columns later may associate an old result with a different column or make the table skip it.

Example: 7
Example:
statistics_result_config['dependent_variable_column'][0]['dependent_variable_column_index'] = ''
Parameter ID: 19
before-after 1st column index
before-after 1st column index text

Zero-based index of the project column used in the before role. Because the interface resolves the column name and options from this index, reordering project columns later may associate an old result with a different column or make the table skip it.

Example: 3
Example:
statistics_result_config['before_after_1st_column_index'] = ''
Parameter ID: 20
before-after 2nd column index
before-after 2nd column index text

Zero-based index of the project column used in the after role. Because the interface resolves the column name and options from this index, reordering project columns later may associate an old result with a different column or make the table skip it.

Example: 4
Example:
statistics_result_config['before_after_2nd_column_index'] = ''
Parameter ID: 21
mean
mean float

Stores the arithmetic mean of analyzed numeric data. When populated, it may be displayed in a metric card or result table; the value gains correct context only when interpreted with row count, analysis type, and related distribution measures.

Example: 42.6
Example:
statistics_result_config['mean'] = ''
Parameter ID: 22
median
median float

Stores the median of analyzed numeric data. When populated, it may be displayed in a metric card or result table; the value gains correct context only when interpreted with row count, analysis type, and related distribution measures.

Example: 41.0
Example:
statistics_result_config['median'] = ''
Parameter ID: 23
standart deviation
standart deviation float

Stores the standard deviation of analyzed numeric data. When populated, it may be displayed in a metric card or result table; the value gains correct context only when interpreted with row count, analysis type, and related distribution measures.

Example: 8.4
Example:
statistics_result_config['standart_deviation'] = ''
Parameter ID: 24
cronbach alpha
cronbach alpha float

Stores the Cronbach alpha internal-consistency coefficient of analyzed numeric data. When populated, it may be displayed in a metric card or result table; the value gains correct context only when interpreted with row count, analysis type, and related distribution measures.

Example: 0.87
Example:
statistics_result_config['cronbach_alpha'] = ''
Parameter ID: 25
minimum
minimum float

Stores the minimum value of analyzed numeric data. When populated, it may be displayed in a metric card or result table; the value gains correct context only when interpreted with row count, analysis type, and related distribution measures.

Example: 18
Example:
statistics_result_config['minimum'] = ''
Parameter ID: 26
maximum
maximum float

Stores the maximum value of analyzed numeric data. When populated, it may be displayed in a metric card or result table; the value gains correct context only when interpreted with row count, analysis type, and related distribution measures.

Example: 69
Example:
statistics_result_config['maximum'] = ''
Parameter ID: 27
sum
sum float

Stores the sum of values of analyzed numeric data. When populated, it may be displayed in a metric card or result table; the value gains correct context only when interpreted with row count, analysis type, and related distribution measures.

Example: 5112
Example:
statistics_result_config['sum'] = ''
Parameter ID: 28
frequency json
frequency json textarea

Stores category-to-count mappings for a categorical column as a JSON object. The interface sums counts to calculate the valid-observation denominator and produces count, group percentage, and overall percentage for each category; malformed JSON prevents table generation.

Example: {"Kadın":64,"Erkek":56}
Example:
statistics_result_config['frequency_json'] = ''
Parameter ID: 29
categorical & categorical json
categorical & categorical json textarea

Stores the nested category-to-count matrix of two categorical variables as a JSON object. The interface produces column-percentage, row-percentage, and raw-frequency tables from this structure; outer keys must represent row groups and inner keys must represent column categories.

Example: {"Tedavi":{"Başarılı":42,"Başarısız":18}}
Example:
statistics_result_config['categorical_categorical_json'] = ''
Parameter ID: 30
categorical & numeric json
categorical & numeric json textarea

Stores numeric summaries by categorical group as a JSON object. Each group may contain mean, median, interquartile range, first and third quartiles, standard deviation, minimum, maximum, sum, and count; cells appear empty when field names do not match the expected structure.

Example: {"Tedavi":{"mean":42.6,"count":60}}
Example:
statistics_result_config['categorical_numeric_json'] = ''
Parameter ID: 31
table
table list_parameter

List carrying prebuilt result tables. The interface first attempts to generate tables from raw JSON fields 29 through 31; when none is produced, it uses parameter 33 name and parameter 34 JSON data from this list for backward compatibility.

Example: regression table
Example:
statistics_result_config['table'] = [{'table_name': '', 'table_json': ''}]
Parameter ID: 32
table name
table name textarea

Name of a table-list item. When it matches a translation key, the interface uses the localized title; otherwise it displays the stored text directly.

Example: logistic_regression
Example:
statistics_result_config['table'][0]['table_name'] = ''
Parameter ID: 33
table json
table json select

JSON data carrying columns and rows arrays for a table-list item. With a valid structure, the interface converts headers and rows into a table; malformed JSON or missing arrays prevents the table from being displayed.

Example: {"columns":["term","coefficient"],"rows":[["age",0.18]]}
Example:
statistics_result_config['table'][0]['table_json'] = ''
Parameter ID: 34
interquartile range
interquartile range text

Stores the interquartile range of the numeric distribution. In median display, the interface combines the first and third quartiles as “median (Q1–Q3)”; this summary is not displayed when a value is missing or cannot be converted to a number.

Example: 12
Example:
statistics_result_config['interquartile_range'] = ''
Parameter ID: 35
interquartile range q1
interquartile range q1 text

Stores the first quartile of the numeric distribution. In median display, the interface combines the first and third quartiles as “median (Q1–Q3)”; this summary is not displayed when a value is missing or cannot be converted to a number.

Example: 35
Example:
statistics_result_config['interquartile_range_q1'] = ''
Parameter ID: 36
interquartile range q3
interquartile range q3 text

Stores the third quartile of the numeric distribution. In median display, the interface combines the first and third quartiles as “median (Q1–Q3)”; this summary is not displayed when a value is missing or cannot be converted to a number.

Example: 47
Example:
statistics_result_config['interquartile_range_q3'] = ''
Parameter ID: 37
levene p value
levene p value text

P value of the Levene test evaluating equality of group variances. The interface displays both fields together in test-selection assessment; when the p value is below the significance threshold, the equal-variance assumption is not supported.

Example: 0.218
Example:
statistics_result_config['levene_p_value'] = ''
Parameter ID: 38
levene test statistic
levene test statistic text

Test statistic of the Levene test evaluating equality of group variances. The interface displays both fields together in test-selection assessment; when the p value is below the significance threshold, the equal-variance assumption is not supported.

Example: 1.534
Example:
statistics_result_config['levene_test_statistic'] = ''
Parameter ID: 39
excess kurtosis values
excess kurtosis values text

JSON dictionary carrying the excess-kurtosis value and observation count for each group. The interface reports marked shape departure when absolute skewness exceeds 2 or absolute excess kurtosis exceeds 7; malformed JSON cannot produce diagnostic rows.

Example: {"Tüm veri":{"value":0.41,"count":120}}
Example:
statistics_result_config['excess_kurtosis_values'] = ''
Parameter ID: 40
skewness values
skewness values text

JSON dictionary carrying the skewness value and observation count for each group. The interface reports marked shape departure when absolute skewness exceeds 2 or absolute excess kurtosis exceeds 7; malformed JSON cannot produce diagnostic rows.

Example: {"Tüm veri":{"value":0.28,"count":120}}
Example:
statistics_result_config['skewness_values'] = ''
Parameter ID: 41
shapiro-wilk results
shapiro-wilk results text

JSON dictionary carrying Shapiro–Wilk availability, test statistic, p value, and observation count for each group. For available groups, the interface assesses normality as rejected when the p value is below the significance threshold.

Example: {"Tüm veri":{"available":true,"statistic":0.98,"p_value":0.12,"count":120}}
Example:
statistics_result_config['shapiro_wilk_results'] = ''
Parameter ID: 42
kolmogorov-smirnov results
kolmogorov-smirnov results text

JSON structure carrying Kolmogorov–Smirnov distribution results by group. The interface displays each group’s test statistic, p value, and observation count from the groups object in the distribution-diagnostics table; without the expected groups structure, these test rows remain empty.

Example: {"groups":{"Tüm veri":{"statistic":0.07,"p_value":0.20,"count":120}}}
Example:
statistics_result_config['kolmogorov_smirnov_results'] = ''
Parameter ID: 43
About Wanalyzer Project / Machine Learning
WISDOMERA / WISTATS

Wanalyzer Project / Machine Learning

Configures the machine-learning analysis definition and execution scope.

Role of the definition

This layer stores the machine-learning analysis definition that runs on the prepared dataset of the shared Wanalyzer project. The definition determines model selection, input columns and their combination mode, output-column interpretation, the K-means cluster count, and automatic execution after data preparation. Because column relationships are stored as zero-based indexes, mappings must be verified again when column order changes. When automatic model and automatic combination are used together, the process can generate many results because every model is tried for every non-empty input subset.

Growth effect: When automatic model and automatic combination are selected together, the estimated result count is model count × input-combination count.
status
status select

Stores the active state of the machine-learning definition. Value 1 stores the definition as active, while value 0 stores it as passive. However, the inspected automatic-analysis selector checks only parameter 14, and the manual execution flow does not contain a shared check that uses this field as a mandatory blocker. Therefore, do not treat passive as a security switch that certainly prevents execution; also set parameter 14 to no to disable automatic execution.

Options and usage purposes
0 — passive
Stores the definition as passive.
1 — active
Stores the definition as active.
Example: 1 — aktif / active
Example:
machine_learning_config['status'] = '1'
Parameter ID: 0
wanalyzer project id
wanalyzer project id text

The client identifier that links the definition to the shared Wanalyzer project, column, and dataset context. When the correct identifier is used, the analysis reads the prepared dataset and column metadata of the same project; an incorrect identifier causes the definition to be searched under another project or prevents the required data and column relationships from being established.

Example: wp_01HXYZ…
Example:
machine_learning_config['wanalyzer_project_id'] = ''
Parameter ID: 1
name
name text

The name used to distinguish the machine-learning definition in lists. When the quick-analysis screen creates a record, it automatically builds this name by combining the model usage mode, selected model, input columns, and output column; a descriptive name makes it easier to identify the configuration of each experiment when a project contains many experiments.

Example: Otomatik — Yaş, Gelir → Satın Alma
Example:
machine_learning_config['name'] = ''
Parameter ID: 2
note
note text

The explanation stored for the purpose, data scope, or evaluation notes of the definition. It is not used as an input variable during model training; therefore, use it to tell people why the experiment was created or under which conditions its results should be interpreted.

Example: Son 12 aylık kayıtlarla yeniden başvuru tahmini.
Example:
machine_learning_config['note'] = ''
Parameter ID: 3
model usage type
model usage type select

Determines the set of models to execute. When selected is used, only the model in parameter 5 is trained. When automatic is used, the backend tries all seven model codes; consequently, the possible result count increases by the product of the model count and the input-combination count, and a model that is incompatible with the data may produce a result record containing a warning instead of a successful metric set.

Options and usage purposes
0 — selected
Runs only the model value in parameter 5.
1 — auto
Tries all seven models for every generated input combination.
Example: 1 — otomatik / auto
Example:
machine_learning_config['model_usage_type'] = '1'
Parameter ID: 4
model
model select

Determines the algorithm to train when model usage type is selected. Logistic regression and the current support-vector-machine path perform classification; linear regression performs regression; random forest, decision tree, and k-nearest neighbor become classifiers or regressors according to whether the output column is interpreted as categorical or numeric. K-means clusters the inputs without using the output column in training.

Options and usage purposes
0 — logistic regression
Builds a classification model.
1 — linear regression
Builds a regression model for numeric output.
2 — support vector machine
Builds a classification model in the current implementation.
3 — k means
Clusters inputs without using the output.
4 — random forest
Performs classification or regression according to output type.
5 — decision tree
Performs classification or regression according to output type.
6 — k nearest neighbor
Performs classification or regression according to output type.
Example: 4 — rassal orman / random forest
Example:
machine_learning_config['model'] = '4'
Parameter ID: 5
cluster count
cluster count integer

The requested number of clusters for the K-means model. If left empty, the backend uses 3; during training, the value is limited to at least 2 and at most the number of usable rows. For example, if 10 is requested but only 6 valid rows exist, the model runs with 6 clusters.

Example: 3
Example:
machine_learning_config['cluster_count'] = ''
Parameter ID: 6
input usage type
input usage type select

Determines how selected input columns are distributed across experiments. With all inputs together, every input is used as one feature set. With automatic combination, every non-empty subset from one column through all columns is generated; because n inputs produce 2ⁿ−1 combinations, execution time and result count can grow quickly. The quick-analysis interface sends this field only when the output is numeric; otherwise, the backend treats an empty value as all inputs together.

Options and usage purposes
0 — all inputs together
Creates one feature set from the selected inputs.
1 — auto combination
Generates every non-empty subset of the selected inputs.
Example: 0 — tüm girdiler birlikte / all inputs together
Example:
machine_learning_config['input_usage_type'] = '0'
Parameter ID: 7
input column
input column list_parameter

The repeating list of columns that the model uses as independent variables or features. At least one input is required. Categorical inputs are converted into indicator columns before training; numeric inputs are converted to numbers, and rows containing invalid numeric values are removed from the training data. The index and usage type in child records are stored separately for each input.

Example: [Yaş, Gelir, Şehir]
Example:
machine_learning_config['input_column'] = [{'input_column_index': '', 'input_column_usage_type': '0'}]
Parameter ID: 8
input column index
input column index text

The zero-based position of the input-list row in the project column array. Value 0 refers to the first column and 1 to the second. If project columns are reordered, the same number points to another column, so the definition must be verified again.

Example: 2
Example:
machine_learning_config['input_column'][0]['input_column_index'] = ''
Parameter ID: 9
input column usage type
input column usage type select

Determines whether each input column is interpreted as categorical or numeric. With automatic, float and integer types in project-column metadata are treated as numeric and all other types as categorical. With categorical, values are expanded into indicator columns; with numeric, values are converted to numbers and rows that cannot be converted are removed. The current quick-analysis record builder stores this field as automatic.

Options and usage purposes
0 — auto
Determines the type from project-column metadata.
1 — categorical
Encodes values as categorical features.
2 — numeric
Converts values into numeric features.
Example: 0 — otomatik / auto
Example:
machine_learning_config['input_column'][0]['input_column_usage_type'] = '0'
Parameter ID: 10
output column index
output column index text

The zero-based position of the dependent variable to predict in the project column array. This column is used as the target during classification and regression training. The K-means backend does not use a target during training; nevertheless, the current quick-analysis interface requires an output column before enabling execution.

Example: 5
Example:
machine_learning_config['output_column_index'] = ''
Parameter ID: 11
output column usage type
output column usage type select

Determines whether the output is interpreted as a categorical or numeric target. With automatic, the project column type is used. A categorical output is converted into class labels; a numeric output is converted to numbers and enables regression models. The current quick-analysis record builder stores this field as automatic.

Options and usage purposes
0 — auto
Determines the type from project-column metadata.
1 — categorical
Interprets the output as a class target.
2 — numeric
Interprets the output as a numeric regression target.
Example: 0 — otomatik / auto
Example:
machine_learning_config['output_column_usage_type'] = '0'
Parameter ID: 12
output column option target category
output column option target category text

The label to treat as the positive class for a categorical output with two options. When supplied, the backend converts rows whose output text exactly matches this label to 1 and all other rows to 0; therefore, the stored label must match the dataset value. The quick-analysis screen displays this field only for an output with exactly two options and stores the visible option label.

Example: Evet
Example:
machine_learning_config['output_column_option_target_category'] = ''
Parameter ID: 13
automatic analysis after project data preparation
automatic analysis after project data preparation select

Determines whether an automatic machine-learning process is created for this definition after the project data-preparation process completes. With yes, the automatic-analysis scan selects the definition and creates a new process record in start status; with no, no automatic process is created after data preparation. Because the inspected selector does not additionally check the status field, the verified way to disable automatic execution is to set this field to no.

Options and usage purposes
0 — no
Does not create an automatic process after data preparation.
1 — yes
Creates a process in start status after data preparation.
Example: 1 — evet / yes
Example:
machine_learning_config['automatic_analysis_after_project_data_preparation'] = '1'
Parameter ID: 14
column filter rule json
column filter rule json textarea

Stores, as JSON, the rule structure generated from column filters on the quick-analysis screen. The interface joins populated rules with an “and” relationship and writes column indexes and values into the record. However, although the inspected MLProcessAnalyzer version parses the field as JSON, it does not call a filtering operation that applies the rule to dataset rows; therefore, in the current backend behavior, populating this field does not guarantee that the rows entering training will change.

Example: {"query_list":[…],"compile_str_array":["(",0,")"]}
Example:
machine_learning_config['column_filter_rule_json'] = {'query_list': [], 'compile_str_array': [], 'compile_str_date': [], 'compile_str_text': []}
Parameter ID: 15
About Wanalyzer Project / Machine Learning / Process
WISDOMERA / WANALYZER

Wanalyzer Project / Machine Learning / Process

Tracks manual and automatic machine-learning executions.

Process flow

The process is created with start status, current start and last-information times, and zero counters. Start and continue are actively tracked; cancel and complete end tracking and refresh results.

Layer separation: Model, calibration interpretation, and result-text language are stored in the process-result child layer.
status
status select

Lifecycle status of the machine-learning analysis process. The record is created with start for both manual and automatic triggers. When start or continue is selected, the interface treats the process as active and refreshes it every five seconds; when cancel or complete is selected, tracking stops and process results are reloaded.

Options and usage purposes
0 — cancel
Removes the process from active tracking and causes results to be refreshed.
1 — start
A new manual or automatic process is created with this value and tracked as active.
2 — continue
Indicates that execution is in progress and continues five-second tracking.
3 — complete
Indicates that execution has finished, stops tracking, and refreshes results.
Example: 1 — başla / start
Example:
machine_learning_process_config['status'] = '1'
Parameter ID: 0
wanalyzer project id
wanalyzer project id text

Client identifier connecting the process to its owning Wanalyzer project. With the correct identifier, the process is stored in the hierarchy of project, machine-learning definition, and process; an incorrect identifier may cause results to be searched under another project context and break the authorization relation.

Example: wp_01HXYZ…
Example:
machine_learning_process_config['wanalyzer_project_id'] = ''
Parameter ID: 1
machine learning id
machine learning id text

Client identifier of the machine-learning definition that produced the process. Because the process list reads child records under this identifier, the correct value makes the execution appear in the history of the related model configuration; an incorrect value associates it with a different definition.

Example: ml_01HXYZ…
Example:
machine_learning_process_config['machine_learning_id'] = ''
Parameter ID: 2
start date
start date datetime

Start time at which the process is created. Both manual and automatic creation code fill this field in Coordinated Universal Time with second precision; if the end date is also recorded, the two values provide the approximate execution duration.

Example: 2026-10-25T09:30:00
Example:
machine_learning_process_config['start_date'] = ''
Parameter ID: 3
end date
end date datetime

Time at which machine-learning execution ends. When populated upon completion, it defines the execution interval together with the start date; it may remain empty while the process is running or when the executor ends without writing a date.

Example: 2026-10-25T09:33:24
Example:
machine_learning_process_config['end_date'] = ''
Parameter ID: 4
last information date
last information date datetime

Shows when the process most recently reported progress. It is initialized to the same value as the start date; when the executor updates it during progress, it becomes an activity indicator, while a value unchanged for a long time may indicate a delayed or stalled process.

Example: 2026-10-25T09:31:10
Example:
machine_learning_process_config['last_information_date'] = ''
Parameter ID: 5
stop date
stop date datetime

Stores the time at which the process was stopped or cancelled. When populated together with cancel status, it explains when execution was terminated; a value in this field alone does not end interface tracking because the tracking decision is based on status.

Example: 2026-10-25T09:31:42
Example:
machine_learning_process_config['stop_date'] = ''
Parameter ID: 6
row count
row count integer

Holds the number of dataset rows associated with or processed by the machine-learning execution. It is initialized to zero when the process is created and indicates training or analysis scope as the executor updates it. Zero alone does not mean failure because the process may have only just started.

Example: 2500
Example:
machine_learning_process_config['row_count'] = ''
Parameter ID: 7
result total count
result total count integer

Holds the total number of machine-learning result records produced under the process. It is initialized to zero; when the executor updates it as results are written to the child layer, it indicates expected result volume. Detailed model, metric, and interpretation data reside in the result child layer, not in this counter.

Example: 1
Example:
machine_learning_process_config['result_total_count'] = ''
Parameter ID: 8
log
log textarea

Text carrying the trigger mode, progress notes, and error descriptions of the process. Manual triggering records authorization or credit execution mode, while automatic triggering records the post-project-data-preparation automatic-analysis note; the log therefore distinguishes why and how the process was started.

Example: Automatic machine learning analysis after project data preparation
Example:
machine_learning_process_config['log'] = ''
Parameter ID: 9
About Wanalyzer Project / Machine Learning / Process / Result
WISDOMERA / WANALYZER

Wanalyzer Project / Machine Learning / Process / Result

Stores the trained model, performance measures, tables, and interpretations.

Result flow

Model type determines interpretation of related classification, regression, or clustering measures. Cross-validation describes generalization stability, while calibration describes probability reliability.

Security and dependency: Pickle must be loaded only from trusted system output. Column indexes depend on project order and table views depend on valid JSON structure.
status
status select

Status field of the result record. Because no status options were supplied and the inspected result view does not filter by this field, no activation behavior has been assumed for particular values; the field carries status information written by the result producer.

Example: completed
Example:
machine_learning_result_config['status'] = ''
Parameter ID: 0
wanalyzer project id
wanalyzer project id text

Client identifier connecting the result to the related Wanalyzer project. With the correct identifier, the result is found in the project → machine learning → process → result hierarchy; an incorrect identifier may direct listing, authorization, or trained-model lookup to a different context.

Example: wp_01HXYZ…
Example:
machine_learning_result_config['wanalyzer_project_id'] = ''
Parameter ID: 1
machine learning id
machine learning id text

Client identifier connecting the result to the related machine-learning definition. With the correct identifier, the result is found in the project → machine learning → process → result hierarchy; an incorrect identifier may direct listing, authorization, or trained-model lookup to a different context.

Example: ml_01HXYZ…
Example:
machine_learning_result_config['machine_learning_id'] = ''
Parameter ID: 2
process id
process id text

Client identifier connecting the result to the related machine-learning process. With the correct identifier, the result is found in the project → machine learning → process → result hierarchy; an incorrect identifier may direct listing, authorization, or trained-model lookup to a different context.

Example: process_01HXYZ…
Example:
machine_learning_result_config['process_id'] = ''
Parameter ID: 3
start date
start date datetime

Time when model-result generation started. Together with the end date, it supports approximate calculation of training and evaluation duration.

Example: 2026-10-25T09:30:00
Example:
machine_learning_result_config['start_date'] = ''
Parameter ID: 4
end date
end date datetime

Time when model-result generation finished. Together with the start date, it defines the execution interval; if empty, generation may be unfinished or the date may not have been written.

Example: 2026-10-25T09:34:12
Example:
machine_learning_result_config['end_date'] = ''
Parameter ID: 5
model
model select

Model family that produced the result. This selection determines the interpretation context for classification, regression, or clustering metrics; for example, accuracy applies to classification, R² and error measures to regression, and silhouette, Calinski–Harabasz, and Davies–Bouldin measures to clustering.

Options and usage purposes
0 — logistic regression
Indicates that the result was produced with this model family.
1 — linear regression
Indicates that the result was produced with this model family.
2 — support vector machine
Indicates that the result was produced with this model family.
3 — k means
Indicates that the result was produced with this model family.
4 — random forest
Indicates that the result was produced with this model family.
5 — decision tree
Indicates that the result was produced with this model family.
6 — k nearest neighbor
Indicates that the result was produced with this model family.
Example: 0 — logistic regression
Example:
machine_learning_result_config['model'] = '0'
Parameter ID: 6
input
input list_parameter

List carrying project columns used as model inputs. Each item contains parameter 8, a zero-based column index; when column order changes, mappings must be checked again because a training variable may resolve to the wrong column name.

Example: age, income
Example:
machine_learning_result_config['input'] = [{'input_column_index': ''}]
Parameter ID: 7
input column index
input column index integer

Zero-based index of the project column used in the model’s input role. The interface resolves the column name from this index in the history title; after reordering, an old result may be displayed with a different column name.

Example: 2
Example:
machine_learning_result_config['input'][0]['input_column_index'] = ''
Parameter ID: 8
output column index
output column index integer

Zero-based index of the project column used in the model’s output role. The interface resolves the column name from this index in the history title; after reordering, an old result may be displayed with a different column name. Unsupervised clustering models may not have a defined output column.

Example: 7
Example:
machine_learning_result_config['output_column_index'] = ''
Parameter ID: 9
model pickle
model pickle textarea

Serialized pickle data of the trained model object. Prediction code attempts to decode this field first as Base64, then as a byte literal, and finally as Latin-1-compatible text before loading it with pickle; when loading fails, no trained model is found. Pickle can execute code from an untrusted source, so only trusted system-generated data must be stored.

Example: gASV…
Example:
machine_learning_result_config['model_pickle'] = ''
Parameter ID: 10
test data ratio
test data ratio float

Ratio of data reserved for evaluation to the complete dataset. A larger ratio bases test metrics on more observations but leaves less data for training; a smaller ratio trains the model on more data but may increase evaluation uncertainty.

Example: 0.20
Example:
machine_learning_result_config['test_data_ratio'] = ''
Parameter ID: 11
accuracy
accuracy float

Proportion of correct predictions among all predictions in classification. With imbalanced classes, high accuracy may hide weak minority-class performance, so it must be interpreted with precision, recall, F-score, and the confusion matrix.

Example: 0.91
Example:
machine_learning_result_config['accuracy'] = ''
Parameter ID: 12
roc auc
roc auc float

Area under the receiver operating characteristic curve summarizing the classification model’s ability to separate classes across thresholds. A higher value indicates stronger overall discrimination, while class balance, calibration, and the operating threshold still require separate evaluation.

Example: 0.94
Example:
machine_learning_result_config['roc_auc'] = ''
Parameter ID: 13
precision (binary)
precision (binary) float

precision value calculated for the positive class of binary classification. A larger value indicates better performance under that evaluation mode; because the averaging method changes the result according to class balance, it must not be confused with the other variants.

Example: 0.90
Example:
machine_learning_result_config['precision_binary'] = ''
Parameter ID: 14
precision (micro)
precision (micro) float

precision value calculated for decisions from all classes through global totals. A larger value indicates better performance under that evaluation mode; because the averaging method changes the result according to class balance, it must not be confused with the other variants.

Example: 0.91
Example:
machine_learning_result_config['precision_micro'] = ''
Parameter ID: 15
precision (macro)
precision (macro) float

precision value calculated for each class with equal weight. A larger value indicates better performance under that evaluation mode; because the averaging method changes the result according to class balance, it must not be confused with the other variants.

Example: 0.89
Example:
machine_learning_result_config['precision_macro'] = ''
Parameter ID: 16
precision (weighted)
precision (weighted) float

precision value calculated for each class weighted by its observation count. A larger value indicates better performance under that evaluation mode; because the averaging method changes the result according to class balance, it must not be confused with the other variants.

Example: 0.90
Example:
machine_learning_result_config['precision_weighted'] = ''
Parameter ID: 17
recall (binary)
recall (binary) float

recall value calculated for the positive class of binary classification. A larger value indicates better performance under that evaluation mode; because the averaging method changes the result according to class balance, it must not be confused with the other variants.

Example: 0.88
Example:
machine_learning_result_config['recall_binary'] = ''
Parameter ID: 18
recall (micro)
recall (micro) float

recall value calculated for decisions from all classes through global totals. A larger value indicates better performance under that evaluation mode; because the averaging method changes the result according to class balance, it must not be confused with the other variants.

Example: 0.91
Example:
machine_learning_result_config['recall_micro'] = ''
Parameter ID: 19
recall (macro)
recall (macro) float

recall value calculated for each class with equal weight. A larger value indicates better performance under that evaluation mode; because the averaging method changes the result according to class balance, it must not be confused with the other variants.

Example: 0.87
Example:
machine_learning_result_config['recall_macro'] = ''
Parameter ID: 20
recall (weighted)
recall (weighted) float

recall value calculated for each class weighted by its observation count. A larger value indicates better performance under that evaluation mode; because the averaging method changes the result according to class balance, it must not be confused with the other variants.

Example: 0.89
Example:
machine_learning_result_config['recall_weighted'] = ''
Parameter ID: 21
f-score (binary)
f-score (binary) float

F-score value calculated for the positive class of binary classification. A larger value indicates better performance under that evaluation mode; because the averaging method changes the result according to class balance, it must not be confused with the other variants.

Example: 0.89
Example:
machine_learning_result_config['f_score_binary'] = ''
Parameter ID: 22
f-score (micro)
f-score (micro) float

F-score value calculated for decisions from all classes through global totals. A larger value indicates better performance under that evaluation mode; because the averaging method changes the result according to class balance, it must not be confused with the other variants.

Example: 0.91
Example:
machine_learning_result_config['f_score_micro'] = ''
Parameter ID: 23
f-score (macro)
f-score (macro) float

F-score value calculated for each class with equal weight. A larger value indicates better performance under that evaluation mode; because the averaging method changes the result according to class balance, it must not be confused with the other variants.

Example: 0.88
Example:
machine_learning_result_config['f_score_macro'] = ''
Parameter ID: 24
f-score (weighted)
f-score (weighted) float

F-score value calculated for each class weighted by its observation count. A larger value indicates better performance under that evaluation mode; because the averaging method changes the result according to class balance, it must not be confused with the other variants.

Example: 0.89
Example:
machine_learning_result_config['f_score_weighted'] = ''
Parameter ID: 25
r2 score
r2 score float

R² score showing how much variance in the target variable is explained by the regression model. Fit improves as the value approaches 1; a value near zero indicates no improvement over predicting the mean, while a negative value may indicate worse performance.

Example: 0.82
Example:
machine_learning_result_config['r2_score'] = ''
Parameter ID: 26
cross-validation method
cross-validation method text

Name of the splitting method used for cross-validation. A method appropriate to the data structure provides a more reliable estimate of generalization; ignoring time, group, or class structure may cause data leakage.

Example: StratifiedKFold
Example:
machine_learning_result_config['cross_validation_method'] = ''
Parameter ID: 27
cross-validation fold count
cross-validation fold count integer

Number of folds into which data is divided for cross-validation. More folds create more training-validation repetitions and increase execution cost; the value must not exceed the number of samples.

Example: 5
Example:
machine_learning_result_config['cross_validation_fold_count'] = ''
Parameter ID: 28
cross-validation accuracy scores
cross-validation accuracy scores text

Text carrying accuracy values calculated for every cross-validation fold. Individual scores expose variation between folds in addition to the mean; the number of values should agree with the fold count.

Example: [0.89, 0.91, 0.90, 0.92, 0.88]
Example:
machine_learning_result_config['cross_validation_accuracy_scores'] = ''
Parameter ID: 29
cross-validation mean accuracy
cross-validation mean accuracy float

Cross-validation mean accuracy obtained from cross-validation repetitions. Unlike a single test split, it summarizes generalization behavior across folds; the mean describes performance, standard deviation describes stability, and minimum and maximum identify the worst and best folds.

Example: 0.90
Example:
machine_learning_result_config['cross_validation_mean_accuracy'] = ''
Parameter ID: 30
cross-validation accuracy standard deviation
cross-validation accuracy standard deviation float

Cross-validation accuracy standard deviation obtained from cross-validation repetitions. Unlike a single test split, it summarizes generalization behavior across folds; the mean describes performance, standard deviation describes stability, and minimum and maximum identify the worst and best folds.

Example: 0.014
Example:
machine_learning_result_config['cross_validation_accuracy_standard_deviation'] = ''
Parameter ID: 31
cross-validation maximum accuracy
cross-validation maximum accuracy float

Cross-validation maximum accuracy obtained from cross-validation repetitions. Unlike a single test split, it summarizes generalization behavior across folds; the mean describes performance, standard deviation describes stability, and minimum and maximum identify the worst and best folds.

Example: 0.92
Example:
machine_learning_result_config['cross_validation_maximum_accuracy'] = ''
Parameter ID: 32
cross-validation minimum accuracy
cross-validation minimum accuracy float

Cross-validation minimum accuracy obtained from cross-validation repetitions. Unlike a single test split, it summarizes generalization behavior across folds; the mean describes performance, standard deviation describes stability, and minimum and maximum identify the worst and best folds.

Example: 0.88
Example:
machine_learning_result_config['cross_validation_minimum_accuracy'] = ''
Parameter ID: 33
cross-validation mean roc auc
cross-validation mean roc auc float

Cross-validation mean roc auc obtained from cross-validation repetitions. Unlike a single test split, it summarizes generalization behavior across folds; the mean describes performance, standard deviation describes stability, and minimum and maximum identify the worst and best folds.

Example: 0.93
Example:
machine_learning_result_config['cross_validation_mean_roc_auc'] = ''
Parameter ID: 34
cross-validation roc auc standard deviation
cross-validation roc auc standard deviation float

Cross-validation roc auc standard deviation obtained from cross-validation repetitions. Unlike a single test split, it summarizes generalization behavior across folds; the mean describes performance, standard deviation describes stability, and minimum and maximum identify the worst and best folds.

Example: 0.012
Example:
machine_learning_result_config['cross_validation_roc_auc_standard_deviation'] = ''
Parameter ID: 35
cross-validation mean r2
cross-validation mean r2 float

Cross-validation mean r2 obtained from cross-validation repetitions. Unlike a single test split, it summarizes generalization behavior across folds; the mean describes performance, standard deviation describes stability, and minimum and maximum identify the worst and best folds.

Example: 0.80
Example:
machine_learning_result_config['cross_validation_mean_r2'] = ''
Parameter ID: 36
cross-validation r2 standard deviation
cross-validation r2 standard deviation float

Cross-validation r2 standard deviation obtained from cross-validation repetitions. Unlike a single test split, it summarizes generalization behavior across folds; the mean describes performance, standard deviation describes stability, and minimum and maximum identify the worst and best folds.

Example: 0.03
Example:
machine_learning_result_config['cross_validation_r2_standard_deviation'] = ''
Parameter ID: 37
cross-validation maximum r2
cross-validation maximum r2 float

Cross-validation maximum r2 obtained from cross-validation repetitions. Unlike a single test split, it summarizes generalization behavior across folds; the mean describes performance, standard deviation describes stability, and minimum and maximum identify the worst and best folds.

Example: 0.84
Example:
machine_learning_result_config['cross_validation_maximum_r2'] = ''
Parameter ID: 38
cross-validation minimum r2
cross-validation minimum r2 float

Cross-validation minimum r2 obtained from cross-validation repetitions. Unlike a single test split, it summarizes generalization behavior across folds; the mean describes performance, standard deviation describes stability, and minimum and maximum identify the worst and best folds.

Example: 0.76
Example:
machine_learning_result_config['cross_validation_minimum_r2'] = ''
Parameter ID: 39
calibration brier score
calibration brier score float

Brier score measuring squared distance between predicted probabilities and observed outcomes. A lower value indicates better probability calibration; it should be evaluated together with class prevalence and model discrimination.

Example: 0.08
Example:
machine_learning_result_config['calibration_brier_score'] = ''
Parameter ID: 40
calibration interpretation
calibration interpretation select

Readable calibration class derived from the Brier score and analysis rules. Good, moderate, or poor summarizes probability reliability; without the numeric Brier score it should not be used alone for definitive model comparison.

Options and usage purposes
0 — good
Classifies calibration as good.
1 — moderate
Classifies calibration as moderate.
2 — poor
Classifies calibration as poor.
Example: 0 — iyi / good
Example:
machine_learning_result_config['calibration_interpretation'] = '0'
Parameter ID: 41
silhouette score
silhouette score float

silhouette score of the clustering result. In general, a higher value indicates that clusters are compact internally and separated from one another; comparisons across data scales and cluster counts must use the same preparation conditions.

Example: 0.61
Example:
machine_learning_result_config['silhouette_score'] = ''
Parameter ID: 42
calinski-harabasz score
calinski-harabasz score float

Calinski–Harabasz score of the clustering result. In general, a higher value indicates that between-cluster separation is strong relative to within-cluster dispersion; comparisons across data scales and cluster counts must use the same preparation conditions.

Example: 420.5
Example:
machine_learning_result_config['calinski_harabasz_score'] = ''
Parameter ID: 43
davies-bouldin score
davies-bouldin score float

Davies–Bouldin score of the clustering result. In general, a lower value indicates that clusters are better separated; comparisons across data scales and cluster counts must use the same preparation conditions.

Example: 0.72
Example:
machine_learning_result_config['davies_bouldin_score'] = ''
Parameter ID: 44
mean absolute error
mean absolute error float

mean absolute error of regression predictions. A lower value indicates better fit and expresses absolute error magnitude in the original target unit; models with different target scales must not be compared directly.

Example: 3.2
Example:
machine_learning_result_config['mean_absolute_error'] = ''
Parameter ID: 45
mean squared error
mean squared error float

mean squared error of regression predictions. A lower value indicates better fit and expresses large errors more strongly by squaring them; models with different target scales must not be compared directly.

Example: 18.4
Example:
machine_learning_result_config['mean_squared_error'] = ''
Parameter ID: 46
root mean squared error
root mean squared error float

root mean squared error of regression predictions. A lower value indicates better fit and expresses large errors strongly while returning the result to the target unit; models with different target scales must not be compared directly.

Example: 4.29
Example:
machine_learning_result_config['root_mean_squared_error'] = ''
Parameter ID: 47
feature importance
feature importance text

Text carrying contributions of input variables to model prediction. Because values may be model-specific coefficients or importance measures, magnitude and sign must be interpreted with the related model method and must not be treated as evidence of causality.

Example: {"age":0.42,"income":0.31}
Example:
machine_learning_result_config['feature_importance'] = ''
Parameter ID: 48
warning
warning text

Warning describing a data, model, or metric limitation encountered during training or evaluation. When populated, the interface displays a prominent warning block; its effect must be evaluated before the result is used.

Example: Azınlık sınıfında örnek sayısı düşüktür.
Example:
machine_learning_result_config['warning'] = ''
Parameter ID: 49
confusion matrix json
confusion matrix json textarea

Confusion-matrix JSON data carrying actual and predicted class counts for classification. The interface processes a matrix or values array with optional labels, columns, and row_labels; without labels it generates zero-based positions, and with an invalid matrix it creates no view.

Example: {"matrix":[[52,4],[7,57]],"labels":["No","Yes"]}
Example:
machine_learning_result_config['confusion_matrix_json'] = ''
Parameter ID: 50
table
table list_parameter

List of table definitions displayed with the result. Each item carries parameter 52 title and parameter 53 JSON data; when valid columns and rows exist, the interface displays the table and allows it to be added to article content.

Example: Model coefficients
Example:
machine_learning_result_config['table'] = [{'table_name': '', 'table_json': ''}]
Parameter ID: 51
table name
table name textarea

User-visible title of a result table. When empty, the interface creates a fallback title from the table parameter name and sequence number; a descriptive title distinguishes different tables.

Example: Model Katsayıları
Example:
machine_learning_result_config['table'][0]['table_name'] = ''
Parameter ID: 52
table json
table json select

JSON value carrying table data structure. The interface accepts an object with columns and rows, an array of objects, or a two-dimensional array whose first row is the header; when none of these forms is provided, the table is not displayed.

Example: {"columns":["variable","coefficient"],"rows":[["age",0.42]]}
Example:
machine_learning_result_config['table'][0]['table_json'] = ''
Parameter ID: 53
result text
result text list_parameter

List carrying readable interpretations of the same result in different languages. Each item stores parameter 55 content with parameter 56 language; the interface selects text matching the current language and displays the first item when no match exists.

Example: Türkçe ve İngilizce yorum
Example:
machine_learning_result_config['result_text'] = [{'result_text_content': '', 'result_text_language': '0'}]
Parameter ID: 54
result text content
result text content textarea

Readable interpretation content in a result-text item. When it explains model performance, limitations, and usage conditions consistently with the metrics, the user can evaluate raw values in the correct context.

Example: Model test verisinde yüksek ayırt etme başarısı gösterdi.
Example:
machine_learning_result_config['result_text'][0]['result_text_content'] = ''
Parameter ID: 55
result text language
result text language select

Turkish or English language code of a result-text item. With the correct code, the interface displays content matching the user’s language; an incorrect code makes the text appear under another language option.

Options and usage purposes
0 — tr
Indicates that the text is the Turkish version.
1 — en
Indicates that the text is the English version.
Example: 0 — tr
Example:
machine_learning_result_config['result_text'][0]['result_text_language'] = '0'
Parameter ID: 56