K-means对地震数据进行处理

首先是……自用

用Spark来对地震数据进行分析，经过预处理后得到各地地震数据，用这些数据进行数据分析，首先是用K-means来求地震聚类

from  pyspark.ml.clustering  import  KMeans
kmeans=KMeans(k=17, seed=2147483648)

from pyspark.ml.feature import VectorAssembler

VectorAssembler =VectorAssembler(inputCols=["Latitude","Longitude","Year","Month", "Day"],outputCol="features")
new_df=VectorAssembler.transform(df)
print(new_df.head(1))
model=kmeans.fit(new_df)    #注意，传入的DataFrame是矢量名称为features的集合
centers=model.clusterCenters()  #产生聚类集合
print(centers)

transformed=model.transform(new_df)
print(transformed.head(1))
dataK = transformed.toPandas()

可视化部分

figK = px.scatter_geo(dataK,
                      color = dataK.prediction,
                      color_continuous_scale = px.colors.sequential.Inferno,
                      lon = dataK.Longitude,