本节介绍用于管理HDC图和HDC算法的方法。请注意，这些方法要求数据库部署HDC服务器。
HDC图
showHDCGraph()
获取从图创建的全部HDC图。
参数
config: RequestConfig（可选）：请求配置。
返回
List[HDCGraph]：获取的HDC图列表。
# Retrieves all HDC graphs of the graph 'miniCircle'
requestConfig = RequestConfig(graph="miniCircle")
hdcGraphs = Conn.showHDCGraph(requestConfig)
for hdcGraph in hdcGraphs:
print(hdcGraph.name, "on", hdcGraph.hdcServerName)
miniCircle_hdc_graph on hdc-server-1
miniCircle_hdc_graph2 on hdc-server-2
createHDCGraphBySchema()
为图创建一个HDC图。
参数
builder: HDCBuilder: 要创建的HDC图；属性hdcGraphName和hdcServerName必填，nodeSchema、edgeSchema和syncType可选。config: RequestConfig（可选）：请求配置。
返回
JobResponse：请求结果。
# Creates an HDC graph named 'test_hdc_graph' for the graph 'miniCircle'
requestConfig = RequestConfig(graph="miniCircle")
hdcBuilder = HDCBuilder(
hdcGraphName="test_hdc_graph",
hdcServerName="hdc-server-1",
nodeSchema=[{"*": ["*"]}],
edgeSchema=[
{"direct": ["*"]},
{"review": ["value", "content"]}
],
syncType=HDCSyncType.SCTATIC
)
response = Conn.createHDCGraphBySchema(hdcBuilder, requestConfig)
jobID = response.jobId
time.sleep(3)
jobs = Conn.showJob(jobID, requestConfig)
for job in jobs:
print(job.id, "-", job.status)
61 - FINISHED
61_1 - FINISHED
dropHDCGraph()
删除指定的HDC图。
参数
hdcGraphName: str：HDC图名称。config: RequestConfig（可选）：请求配置。
返回
Response：请求结果。
# Drops the HDC graph 'miniCircle_hdc_graph2' of the graph 'miniCircle'
requestConfig = RequestConfig(graph="miniCircle")
response = Conn.dropHDCGraph("miniCircle_hdc_graph2", requestConfig)
print(response.status.code.name)
SUCCESS
HDC算法
showHDCAlgo()
获取安装在一台HDC服务器上的全部HDC算法。
参数
hdcServerName: str：HDC服务器名称。config: RequestConfig（可选）：请求配置。
返回值
List[Algo]：获取的HDC算法列表。
# Retrieves all HDC algorithms installed on the HDC server 'hdc-server-1'
algos = Conn.showHDCAlgo("hdc-server-1")
data = [[algo.name, algo.writeSupportType] for algo in algos if algo.type == "algo"]
headers = ["Name", "writeSupportType"]
print(tabulate.tabulate(data, headers=headers, tablefmt="grid"))
+-----------+--------------------+
| Name | writeSupportType |
+===========+====================+
| fastRP | DB,FILE |
+-----------+--------------------+
| struc2vec | DB,FILE |
+-----------+--------------------+
installHDCAlgo()
在一台HDC服务器上安装一个HDC算法。
参数
files: List[str]：安装文件路径列表，算法包文件（.so）必须，配置文件（.yml）可选。hdcServerName: str：HDC服务器名称。config: RequestConfig（可选）：请求配置。
返回值
Response：请求结果。
# Installs the HDC algorithm LPA on the HDC server 'hdc-server-1'
# The files 'libplugin_lpa.so' and 'lpa.yml' are located in the 'algo' folder that is placed in the same directory as the file you executed
response = Conn.installHDCAlgo(["algo/libplugin_lpa.so", "algo/lpa.yml"], "hdc-server-1")
print(response.status.code.name)
SUCCESS
uninstallHDCAlgo()
从一台HDC服务器上卸载一个HDC算法。
参数
algoName: str：算法名称。hdcServerName: str：HDC服务器名称。config: RequestConfig（可选）：请求配置。
返回值
Response：请求结果。
# Uninstalls the HDC algorithm LPA from the HDC server 'hdc-server-1'
response = Conn.uninstallHDCAlgo("lpa", "hdc-server-1")
print(response.status.code.name)
SUCCESS
rollbackHDCAlgo()
将一台HDC服务器上的一个指定HDC算法回退到上一个版本。
参数
algoName: str：算法名称。hdcServerName: str：HDC服务器名称。config: RequestConfig（可选）：请求配置。
返回值
Response：请求结果。
# Rolls back the HDC algorithms LPA on the HDC server 'hdc-server-1'
response = Conn.rollbackHDCAlgo("lpa", "hdc-server-1")
print(response.status.code.name)
SUCCESS
完整示例
from ultipa import UltipaConfig, Connection
ultipaConfig = UltipaConfig()
# URI example: ultipaConfig.hosts = ["https://mqj4zouys.us-east-1.cloud.ultipa.com:60010"]
ultipaConfig.hosts = ["192.168.1.85:60061", "192.168.1.87:60061", "192.168.1.88:60061"]
ultipaConfig.username = "<username>"
ultipaConfig.password = "<password>"
Conn = Connection.NewConnection(defaultConfig=ultipaConfig)
# Installs the HDC algorithm LPA on the HDC server 'hdc-server-1'
# The files 'libplugin_lpa.so' and 'lpa.yml' are located in the 'algo' folder that is placed in the same directory as the file you executed
response = Conn.installHDCAlgo(["algo/libplugin_lpa.so", "algo/lpa.yml"], "hdc-server-1")
print(response.status.code.name)