Run and return value of class function inside __init__ Python Class Constructor

Question

Hello Stackoverflow folks,... I hope this questions is not already answered. After half a day of googeling I did resign myself to asking a question here. My problem is the following:

I want to create a class which takes some information and processes this information:

 #Klassendefinition für eine Instanz von Rohdaten
class raw_data(): 
 def __init__(self, filename_rawdata, filename_metadata,
 file_format, path, category, df_raw, df_meta):
 self.filename_rawdata = filename_rawdata
 self.filename_metadata = filename_metadata
 self.file_format = file_format
 self.path = path
 self.category = category
 self.df_raw = getDF(self.filename_rawdata)
 self.df_meta = getDF(self.filename_metadata)
 # generator
 def parse(self, path):
 g = gzip.open(path, 'rb')
 for l in g:
 yield eval(l)
 # function that returns a pandas dataframe with the data
 def getDF(self, filename):
 i = 0
 df = {}
 for d in self.parse(filename):
 df[i] = d
 i += 1
 return pd.DataFrame.from_dict(df, orient='index')

Now I have a problem with the init method, I would like to run the class method below on default when the class in instantiated, but I somehow cannot manage to get this working. I have seen several other posts here like [Calling a class function inside of __init__ [1]: Python 3: Calling a class function inside of __init__ but I am still not able to do it. The first question did work for me, but I would like to call the instance variable after the constructor ran.

I tried this:

class raw_data(): 
 def __init__(self, filename_rawdata, filename_metadata,
 file_format, path, category):
 self.filename_rawdata = filename_rawdata
 self.filename_metadata = filename_metadata
 self.file_format = file_format
 self.path = path
 self.category = category
 getDF(self.filename_rawdata)
 getDF(self.filename_metadata)
 # generator
 def parse(self, path):
 g = gzip.open(path, 'rb')
 for l in g:
 yield eval(l)
 # function that returns a pandas dataframe with the data
 def getDF(self, filename):
 i = 0
 df = {}
 for d in self.parse(filename):
 df[i] = d
 i += 1
 return pd.DataFrame.from_dict(df, orient='index')

But I get an error because getDF is not defined (obviously).. I hope this questions is not silly by any means. I need to do it that way, because afterwards I want to run like 50-60 instance calls and I do not want to repeat like Instance.getDF() ... for every instance, but rather would like to have it called directly.

• Phydeaux: self.getDF(...)
• chepner: As an aside, raw_data.getDF can be reduced to return pd.DataFrame.from_dict(dict(enumerate(self.parse(filename)))) or (I think) return pd.DataFrame.from_records(enumerate(self.parse(filename))).
• chepner: Also, neither parse nor getDF need to be methods of your class; they could be defined as regular functions outside the class, or, if you really want to keep them in your class's namespace, be made static methods.

Answer 1

All you need to so is call getDF like any other method, using self as the object on which it should be invoked.

self.df_raw = self.getDF(self.filename_rawdata)

That said, this class could be greatly simplified by making it a dataclass.

from dataclasses import dataclass
@dataclass
class RawData:
 filename_rawdata: str
 filename_metadata: str
 path: str
 category: str
 def __post_init__(self):
 self.df_raw = self.getDF(self.filename_rawdata)
 self.df_meta = self.getDF(self.filename_metadata)
 @staticmethod
 def parse(path):
 with gzip.open(path, 'rb') as g:
 yield from map(eval, g)
 @staticmethod
 def getDF(filename):
 return pd.DataFrame.from_records(enumerate(RawData.parse(filename)))

The auto-generated __init__ method will set the four defined attributes for you. __post_init__ will be called after __init__, giving you the opportunity to call getDF on the two given file names.

• BayerischerSchweitzer: Ah shit, I did look at the code x-times but somehow did not see that I am missing the ".self" before the function :-( sorry this was stupid from me. Never heard of this "dataclasses" before, this was an eyeopener for me, thank you so much!!