Question

from pyspark import SparkConf, SparkContext
from rdkit import Chem
from rdkit import DataStructs
from rdkit.Chem import AllChem
from rdkit.Chem import MACCSkeys


def parseLine(line): #get the names from the csv file
    fields = line.split('\t')
    canonical_smiles = fields[1]
    return canonical_smiles

def transform(smile):
    m1 = Chem.MolFromSmiles(smile)
    fp1_4 = AllChem.GetMorganFingerprint(m1,2)
    return fp1_4

#set the configuartion
conf = SparkConf().setMaster("local").setAppName("Similarity")
sc = SparkContext(conf = conf)

lines = sc.textFile("PS22_smiles_only_3.tsv") #loads the data
all_data = lines.map(parseLine) #method

similarity_fcp4 = all_data.map(similarity) #method

results = similarity_fcp4.collect() #action

for result in results:
    print(result)

到目前为止，这是我的代码。它读取一个tsv文件，然后根据ecfp4指纹转换列的每一行。我坚持的是如何将rdd的每个元素与所有其他元素进行比较。在普通的python中，我的代码是一个for循环，

for j in range(i+1,  df.shape[0]):
    fp2_4 = AllChem.GetMorganFingerprint(m2,2)
    sim_ecfp4 = DataStructs.DiceSimilarity(fp1_4,fp2_4)

我如何在pyspark中做到这一点？

计算rdd中每个元素与其所有其他元素之间的相似度

0 个答案: