def main(self, ds):
df = ds.get("titanic").content
df[self.column] = df[self.column].fillna(df[self.column].median())
rds = DataSet()
rds.put(f"fillna_{self.column}", DataFrameData(df[self.column]))
return rds
def main(self, ds):
df = ds.get("titanic").content
df.loc[df["Sex"] == "male", "Sex"] = 0
df.loc[df["Sex"] == "female", "Sex"] = 1
rds = DataSet()
rds.put("sex_to_code", DataFrameData(df["Sex"]))
return rds
def main(self, ds):
repo = LocalFileRepository(
Path(os.path.dirname(__file__)) / Path("../titanic.csv"))
titanic_data = DataFrameData.load(repo)
ds = DataSet()
ds.put("titanic", titanic_data)
return ds
def main(self, ds):
df = ds.get("titanic").content
df.loc[df["Sex"] == "male", "Sex"] = 0
df.loc[df["Sex"] == "female", "Sex"] = 1
ds = DataSet()
ds.put("titanic", DataFrameData(df))
return ds
def main(self, ds):
df = ds.get("titanic").content
df["Embarked"] = df["Embarked"].fillna("S")
df.loc[df["Embarked"] == "S", "Embarked"] = 0
df.loc[df["Embarked"] == "C", "Embarked"] = 1
df.loc[df["Embarked"] == "Q", "Embarked"] = 2
rds = DataSet()
rds.put("embarked_to_code", DataFrameData(df["Embarked"]))
return rds
def main(self, ds):
df = ds.get("titanic").content
df["Sex"][df["Sex"] == "male"] = 0
df["Sex"][df["Sex"] == "female"] = 1
rds = DataSet()
rds.put("sex_to_code", DataFrameData(df["Sex"]))
time.sleep(random.randint(3, 10))
return rds
def main(self, ds):
df = ds.get("titanic").content
df = df.drop(["Age", "Sex", "Embarked"], axis=1)
df = df.join(ds.get("fillna_Age").content)
df = df.join(ds.get("sex_to_code").content)
df = df.join(ds.get("embarked_to_code").content)
rds = DataSet()
rds.put("titanic_result", DataFrameData(df))
return rds
def main(self, ds):
df = ds.get("titanic").content
df["Embarked"] = df["Embarked"].fillna("S")
df["Embarked"][df["Embarked"] == "S"] = 0
df["Embarked"][df["Embarked"] == "C"] = 1
df["Embarked"][df["Embarked"] == "Q"] = 2
rds = DataSet()
rds.put("embarked_to_code", DataFrameData(df["Embarked"]))
time.sleep(random.randint(3, 10))
return rds
df.loc[df["Embarked"] == "Q", "Embarked"] = 2
ds = DataSet()
ds.put("titanic", DataFrameData(df))
return ds
if __name__ == "__main__":
basicConfig(level=DEBUG)
# データセットの読み込み
ds = DataSet()
repo = LocalFileRepository(
Path(os.path.dirname(__file__)) / Path("../titanic.csv"))
titanic_data = DataFrameData.load(repo)
ds.put("titanic", titanic_data)
#
print("## Original data")
print(ds.get("titanic").content)
# Graphで処理する
# Age欠損埋め -> 性別のコード化 -> 乗船した港 のコード化 の順で処理
graph = Graph()
fill_age = graph.append(FillNaMedian("Age"))
sex_to_code = graph.append(SexToCode(), [fill_age])
graph.append(EmbarkedToCode(), [sex_to_code])
ds = graph.run(ds)
print("## Processed data")
print(ds.get("titanic").content)
repo_s = SqlAlchemyRepository(engine)
md = SqlAlchemyModelData(repo_s, Titanic)
md.update_dataframe(titanic_data.content)
md.save()
if __name__ == "__main__":
basicConfig(level=DEBUG)
# データセットの読み込み・DBの準備
engine = create_engine("sqlite:///example.sqlite3", echo=True)
prepare_db(engine)
repo = SqlAlchemyRepository(engine)
d = SqlAlchemyModelData(repo, Titanic)
d.query()
passenger_ids = [m.PassengerId for m in d.content]
# データを一行ずつ SQLAlchemy のモデルに取り出し、処理して書き戻す例
for passenger_id in passenger_ids:
d.query(lambda x: x.filter(Titanic.PassengerId == passenger_id))
ds = DataSet()
ds.put("titanic", d)
ds = SexToCode().main(ds)
ds = EmbarkedToCode().main(ds)
ds.save_all()
