开源一个蛋白质的

包括数据初步统计和一个简单的baseline

1.45229分，目前可以排22名

好久没来DC了

#encoding=utf-8

import pandas as pd

import numpy as np

import sys

df_protein_train = pd.read_csv('./data/df_protein_train.csv')

df_protein_train['seq_len'] = df_protein_train['Sequence'].apply(len)

print "protein_train"

print df_protein_train.head()

print df_protein_train.describe()

df_protein_test = pd.read_csv('./data/df_protein_test.csv')

df_protein_test['seq_len'] = df_protein_test['Sequence'].apply(len)

print "protein_test"

print df_protein_test.head()

print df_protein_test.describe()

df_molecule = pd.read_csv('./data/df_molecule.csv')

df_molecule['Fingerprint_len'] = df_molecule['Fingerprint'].apply(len)

print "molecure"

print df_molecule.head()

print df_molecule.describe()

df_affinity_train = pd.read_csv('./data/df_affinity_train.csv')

print "affinity"

print df_affinity_train.head()

print df_affinity_train.describe()

df_affinity_test = pd.read_csv('./data/df_affinity_test_toBePredicted.csv')

print "affnity toBePredicted"

print df_affinity_test.describe()

df_molecule_avg = df_affinity_train.groupby(['Molecule_ID'],as_index=False)['Ki'].agg({'Ki_avg':'mean'})

print "affinity group by Molecule_ID"

print df_molecule_avg.describe()

df1 = pd.merge(df_affinity_train,df_protein_train, on=["Protein_ID"],how='left')

print "merge df_affinity_train,df_protein_train"

print df1.describe()

df1 = pd.merge(df_affinity_train,df_protein_test, on=["Protein_ID"],how='left')

print "merge df_affinity_train,df_protein_test"

print df1.describe()

df1 = pd.merge(df_affinity_test,df_molecule_avg, on=["Molecule_ID"],how='left')

df1 = df1.fillna(df1.mean()['Ki_avg'])

df1.columns = ['Protein_ID','Molecule_ID','Ki']

df1.to_csv("result.csv",index=False)