SankeyFlow

https://pypi.org/project/sankeyflow/
Admin
SankeyFlow is a lightweight Python package that creates Sankey flow diagrams using Matplotlib with...
Tokens:2,037
Snippets:23
Trust Score:10
Update:5 months ago
Show doc for...
Tokens:
Raw
### Install SankeyFlow Package

Source: https://pypi.org/project/sankeyflow

Use this command to install the SankeyFlow package using pip. Ensure you have Python 3.8 or higher.

```bash
python3 -m pip install sankeyflow

```

--------------------------------

### Configure Flow Color Mode

Source: https://pypi.org/project/sankeyflow

This example illustrates how to control the color of flows based on their source or destination node. The `flow_color_mode` can be set globally or per flow.

```python
flows = [
    ('A', 'B1', 4),
    ('A', 'B2', 5, {'flow_color_mode': 'dest'}),
    ('B1', 'C', 1),
    ('B2', 'C', 2),
]

s = Sankey(
    flows=flows,
    nodes=nodes,
    flow_color_mode='source',
)
s.draw()

```

--------------------------------

### Import Sankey Class

Source: https://pypi.org/project/sankeyflow

Import the Sankey class from the sankeyflow library to start creating diagrams.

```python
from sankeyflow import Sankey

```

--------------------------------

### Create Sankey Diagram with Explicit Nodes and Flows

Source: https://pypi.org/project/sankeyflow

This example shows how to define both nodes and flows explicitly to create a Sankey diagram. Nodes are structured by levels, and flows connect sources to destinations.

```python
nodes = [
    [('A', 10)],
    [('B1', 4), ('B2', 5)],
    [('C', 3)]
]
flows = [
    ('A', 'B1', 4),
    ('A', 'B2', 5),
    ('B1', 'C', 1),
    ('B2', 'C', 2),
]

plt.figure(figsize=(4, 3), dpi=144)
s = Sankey(flows=flows, nodes=nodes)
s.draw()

```

--------------------------------

### Create and Draw a Basic Sankey Diagram

Source: https://pypi.org/project/sankeyflow

This snippet demonstrates the basic usage of SankeyFlow to create and draw a Sankey diagram from a list of flows. It requires Matplotlib for display.

```python
import matplotlib.pyplot as plt
from sankeyflow import Sankey

flows = [
    ('Product', 'Total revenue', 20779),
    ('Sevice and other', 'Total revenue', 30949),
    ('Total revenue', 'Gross margin', 34768),
    ('Total revenue', 'Cost of revenue', 16960),
    ...
]
s = Sankey(flows=flows)
s.draw()
plt.show()

```

--------------------------------

### Configure Inferred Nodes via List

Source: https://pypi.org/project/sankeyflow

Retrieve inferred nodes, edit the list, and then pass it to `Sankey.sankey()`. This method allows for editing node ordering and levels.

```python
nodes = Sankey.infer_nodes(flows)
# edit nodes
s.sankey(flows, nodes)

```

--------------------------------

### Customize Node Label Formatting

Source: https://pypi.org/project/sankeyflow

This code snippet shows how to customize node labels, including their format and position. The `label_format` option allows for dynamic formatting of node labels with values.

```python
nodes = [
    [('A', 10)],
    [('B1', 4), ('B2', 5)],
    [('C', 3, {'label_pos':'top'})]
]
flows = [
    ('A', 'B1', 4),
    ('A', 'B2', 5),
    ('B1', 'C', 1),
    ('B2', 'C', 2),
]

s = Sankey(
    flows=flows,
    nodes=nodes,
    node_opts=dict(label_format='{label} ${value:.2f}'),
)
s.draw()

```

--------------------------------

### Automatic Node Inference with Sankey

Source: https://pypi.org/project/sankeyflow

Demonstrates automatic node inference by setting `nodes=None`. Nodes are ordered as they appear in the flows. Use this when you don't need to pre-define nodes.

```python
gross = [
    ('Gross margin', 'Operating\nincome', 200),
    ('Gross margin', 'MG&A', 100), 
    ('Gross margin', 'R&D', 100), 
]
income = [
    ('Operating\nincome', 'Income', 200),
    ('Other income', 'Income', 100, {'flow_color_mode': 'source'}),
]

plt.subplot(121)
s1 = Sankey(flows=gross + income)
s1.draw()

plt.subplot(122)
s2 = Sankey(flows=gross[:1] + income + gross[1:])
s2.draw()

plt.tight_layout()

```

--------------------------------

### Customize Sankey Diagram Appearance (Color, Curvature)

Source: https://pypi.org/project/sankeyflow

This snippet demonstrates how to customize the Sankey diagram's appearance by setting a colormap, disabling flow curvature, and specifying a flow color.

```python
flows = [
    ('A', 'B1', 4),
    ('A', 'B2', 5),
    ('B1', 'C', 1),
    ('B2', 'C', 2, {'color': 'red'})
]

s = Sankey(
    flows=flows,
    nodes=nodes,
    cmap=plt.cm.Pastel1,
    flow_opts=dict(curvature=0),
)
s.draw()

```

--------------------------------

### Configure Inferred Nodes Directly

Source: https://pypi.org/project/sankeyflow

Access and modify inferred nodes directly after Sankey object creation. This is useful for changing labels of existing nodes.

```python
s = Sankey(flows=flows)
s.find_node('name')[0].label = 'My label'

```
Tokens:
Raw