Use function (not using lambda) with apply method for pandas DataFrame

Question

I am trying to add column according to the existing column in dataframe following this page.
Since there are many if case, I defined the if case as function (getDirection method) and tried to call by apply method(addDirection method).
Howerver, I got following error.

TypeError: getDirection() takes 1 positional argument but 2 were given

Can anyone tell how I can call function by apply for dataframe?
Code is shown below.

def addDirection(self):
    group=self.group
    group['direction']=group['Azimuth [deg]'].apply(self.getDirection)

   def getDirection(angle):
        if angle < 11.25 or angle >= 348.75:
            return "N"
        elif angle < 33.75 and angle >= 11.25:
            return "NNE"
        elif angle < 56.25 and angle >= 33.75:
            return "NE"
        elif angle < 78.75 and angle >= 56.25:
            return "ENE"
        elif angle < 101.25 and angle >= 78.75:
            return "E"
        elif angle < 123.75 and angle >= 101.25:
            return "ESE"
        elif angle < 146.25 and angle >= 123.75:
            return "SE"
        elif angle < 168.75 and angle >= 146.25:
            return "SSE"
        elif angle < 191.25 and angle >= 168.75:
            return "S"
        elif angle < 213.75 and angle >= 191.25:
            return "SSW"
        elif angle < 236.25 and angle >= 213.75:
            return "SW"
        elif angle < 258.75 and angle >= 236.25:
            return "WSW"
        elif angle < 281.25 and angle >= 258.75:
            return "W"
        elif angle < 303.75 and angle >= 281.25:
            return "WNW"
        elif angle < 326.25 and angle >= 303.75:
            return "NW"
        elif angle < 348.75 and angle >= 326.25:
            return "NNW"

getDirection appears to be a method in some class, not a regular function. — chepner
– chepner, Commented Dec 3, 2017 at 14:20
@akilat90 Referring to this page, I suppose below script is applying the getDirection() method to dataframe. 'group['direction']=group['Azimuth [deg]'].apply(self.getDirection)' — Katsuya Obara
– Katsuya Obara, Commented Dec 3, 2017 at 14:20
This is a lot more concisely implemented as a list lookup: return ["N", "NNE", "NE, "ENE", ...][int((angle*100)//1125]. — chepner
– chepner, Commented Dec 3, 2017 at 14:27

Cornflex · Accepted Answer · 2017-12-03 14:49:09Z

When calling self.getDirection(a), you are implicitly calling getDirection(self, a). This is why the error message states that two arguments were given instead of one (self and the element of the series).

The way you call the function (prepending self.) implies that it is not a static method, but rather an instance method. Instance methods are defined inside a class; their result typically depends on the internal state of the object (instance of the class) you are calling it on.

If you called the function as an instance method by mistake, you should write apply(getDirection) instead of apply(self.getDirection). Most likely this is what you want, since the function's outcome depends only on angle and not on any object's internal state.

If you want getDirection to be an instance method, define the function as def getDirection(self, a). You don't have to change anything else.

See also this answer on instance and class methods.

Bharath M Shetty · Accepted Answer · 2017-12-03 15:06:17Z

As @Cornflex said is because you are passing additional paramter using self so you got that error. Either do def getDirection(self,angle) or just apply.(getDirection)

Since its pandas instead of that many if else's we can just do binning using pd.cut i.e

df = pd.DataFrame({'Azimuth [deg]':[300,340,150]})
bins = [11.25, 33.75, 56.25, 78.75, 101.25, 123.75, 146.25, 168.75, 191.25, 213.75, 236.25, 258.75, 281.75, 303.75, 326.25, 348.75]

labels = ['NNE', 'NE', 'ENE', 'E', 'ESE', 'SE', 'SSE', 'S', 'SSW', 'SW', 'WSW', 'W', 'WNW', 'NW', 'NNW']

pd.cut(df['Azimuth [deg]'],bins=bins,labels=labels).fillna('N') # fillna is for your first condition.

0    WNW
1    NNW
2    SSE
Name: Azimuth [deg], dtype: category
Categories (15, object): [NNE < NE < ENE < E ... W < WNW < NW < NNW]
In [557]:

df['Azimuth [deg]'].apply(getDirection)

0    WNW
1    NNW
2    SSE
Name: Azimuth [deg], dtype: object

Collectives™ on Stack Overflow

Use function (not using lambda) with apply method for pandas DataFrame

2 Answers 2

Comments

Comments

Your Answer

Linked

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

Comments

Comments

Your Answer

Sign up or log in

Post as a guest

Linked

Related