Tukukkk

I am working with a Pandas DataFrame of sports scores which contains a Series 'Score'. All items in this Series contain both team's scores in a single string, separated by hyphen, with no spaces, for example

('25-7', '6-2', ...)

I am attempting to split each value into 2 separate lists: left_score and right_score using Jupyter notebook. I have used the str.split('-') method for Series, which is supposed to convert each string into a list such that my scores would be

['25','7'], ['6','2']

However, when I run this it executes, but it does not recognize the hyphen, and returns the entire string as index 0.

I have tried using '-' and "-" with no difference. I also tried using a for loop and using the Python core str.split(). The core function works on a standalone string in Jupyter as expected, but when run in a loop, it again returns the entire string as the only element.

I've tried accessing the strings within the Series directly as well, and the function still fails. The following should return '25', but it returns '25-7'.

dataframe_name.Score.str.split("-").str[0][0]`

Really enjoying working with Pandas and DataFrames, but the syntax is proving a challenge - any thoughts appreciated.

EDIT: Adding sample code as requested. Note this is across multiple Jupyter cells, but I am executing them in sequence.

In[1]:

import numpy as np

import pandas as pd

import matplotlib.pyplot as plt

%matplotlib inline



df = pd.read_csv('./file_name.csv', sep='t')



df.head(3)

Out[1]:

df

_    Score

0    25-7

1    6-2

2    4-4

In[2]:

# Thanks to user Pygo, I attempted the suggested solution to no avail:

df['Score'].str.split('-',n=1,expand=False).values.tolist()

Out[2]:

[['25-7'],

['6-2'],

['4-4'],

... ]

• Jupyter Notebook version 5.5.0


• Anaconda version 5.2.0


• Python version 3.6.5


• Pandas version 0.23.0


• Numpy version 1.14.3

Is it possible there is a version or reference conflict?

EDIT2:

I tried iterating through each letter in the string to perform the split function manually, and have now discovered that .join(), += are not working inside of for loops either. Where would I look for a Pandas and/or Core String malfunction in Jupyter Notebook loops?

We can use the split function to split the Score column at every "-". Then parameter is set to 1 as the maximum number of separations in a single string will be 1. The expand parameter is False(If False, return Series/Index/DataFrame).

Example DataFrame:

df

   Score

0   25-7

1    6-2

2  19-22

Expected result : Using str.split + values.tolist()

df['Score'].str.split('-', n=1, expand=False).values.tolist()

[['25', '7'], ['6', '2'], ['19', '22']]

Hope this will help on the bare minimum information provided.

The Series that I was attempting to parse at the - character was failing at my troubleshooting boolean condition for if letter == '-' ... I realized that the data in my Series had the other kind of hyphen (m-hyphen vs n-hyphen; one is a "wide" character where the other is a "normal" character.) In Jupyter, these look indistinguishable - if there is a trick to discerning these within the notebook, I would love to learn it!