[WIP] Segmentation Tree: segmentation_tree - MaksimAbelchuk/RnS_Docs GitHub Wiki



Overview
visualization_id: segmentation_tree
Configuration guide
Basic info
The structure of the basic info is similar to any other chart:
name: segmentation-tree
title: Segmentation Tree
connection_name: test_chem
dataset_name: test-chem
visualization_id: segmentation_tree
Multiple trees
The feature allows building various charts at a moment in one chart.
chart_config:
multiple: true
Row type
chart_config:
row_type: type
Tree
Layout
chart_config:
tree:
layout:
orientation: horizontal
max_zoom_value: 15
group_padding: 24
siblings_padding: 32
node_levels_padding: 44
Nodes
chart_config:
tree:
node:
value: node_value
type: node_type
depth: node_depth
id: node_id
variable: node_variable
relation: node_relation
condition_value: node_condition_value
Node indicators
chart_config:
tree:
node_indicators:
show_continuous: false
show_categorical: 0
show_info_node: false
Node structure
chart_config:
tree:
node_struct:
id: struct_node_id
leaf_id: struct_leaf_node_id
Variables
chart_config:
tree:
vars:
histogram_stick: recommended_price
histogram_variable: unit_price
impact_baseline: target_value
impact_delta: impact
plain_text_variable: customer_id
plain_text_target_variable: unit_price
tree_selection_variable: product_id
Grid
chart_config:
grid:
page_size: 5
max_column_width: 25
columns:
- data_node_id
- customer_id_orig
- customer_score
- volume
- unit_price
- margin
- revenue_orig
- n_prod
- avg_price_per_prod
- revenue_adjusted
- total_revenue_adj
- product_share
- revenue
- product_id
- customer_id
- region
- payment_terms
- packaging
- transport
- key
- cluster_level
- avg_cust_prod_rev_size_num
- avg_cust_prod_vol_size_num
- distinct_cust_per_prod_num
- total_cust_prod_trans_counts_num
- total_cust_trans_counts_num
- volume_average_num
- predicted_values
- train_flag
- percentile
- target_value
- recommended_target_value
- recommended_price
- impact
- value_at_stake
row:
node_id: data_node_id
- multiple - can be true or false. If true, then you can use multiple scenarios (datasets).
- row_type - define the type of the row.
- scenario - dimension used to distinguish scenarios.
- tree - dimension used to distinguish trees.
- node_indicator
- show_continuous - can be true or false. If true, then the continuous indicator will be shown on the chart.
- show_categorical: 1
- histogram_bars_default - default number of bars in histogram (in each node).
- histogram_bars_variants - number of bars in histogram (Can be 4, 5, 10, 20).
- layout:
- orientation - how the tree is placed. Can be horizontal and vertical.
- group_padding - padding between groups in pixels.
- siblings_padding - padding between nodes in one group in pixels
- node_levels_padding - adding between levels in pixels.
- node_struct:
- id - a dimension where the node is stored
- leaf_id - a dimension where leaf-node id is stored
- vars:
- variable - a dimension where the variable is stored
- value - a dimension where the value is stored
- grid:
- max_column_width - maximal width of a column.
- columns - names which will be listed in the footer of the chart.
- row:
- node_id - a dimension where node id stored
- types:
- other:
-
- PlainTextType
-
- LocalHistogramType
-
- GlobalHistogramType
-
- GlobalPercentileHistogramType
-
- NodeImpactType
-
- TreeImpactType
-
- settings:
- scales_id: numbers
- histogram_bars_default: 10
- histogram_bars_variants:
-
- 4
-
- 5
-
- 10
-
- 20
-
- node_indicator
- row_type - define the type of the row.
Other features
Remove 'Rows' on a leaf node
You can remove 'Rows' from a leaf node by using this property in the config:
chart_config:
plain_text_show_rows: false