how to converting pyspark matrix to dataframe

Convert PySpark CoordinateMatrix into PySpark Dataframe got the CoordinateMatrix from cosine similarity calculation need to convert to PySpark Dataframe so I can convert to pandas DataFrame for analysis

from pyspark.mllib.linalg.distributed import IndexedRow, IndexedRowMatrix

rdd =
mymatrix = IndexedRowMatrix( row: IndexedRow(row[0], row[1:]))).toBlockMatrix().transpose().toIndexedRowMatrix()
mysimilarities = mymatrix.columnSimilarities()

i used this for cosine similarity and the matrix contains coordinates of cosine similarity