I have a dataset with athletes, their age-range and the year they competed:

import ipywidgets as widgets
from ipywidgets import interact  
import pandas as pd  
import matplotlib.pyplot as plt 
import seaborn as sns  
import numpy as np  

data = {'Name':['Tom', 'Nick', 'Chris', 'Jack'],
        'Age':['20-25', '19<', '>75', '20-25'], 
        'Year':[2019,2019,2018,2018], 
         'Count': [1,1,1,1]} 
df = pd.DataFrame(data)

Mind you, this is 1000s on entries long, but I didn't want to use that much space so I'm using a much simplified version.

I am trying to make an interactive bar plot with ipywidgets using that data.

age = ['20-25', '19<', '>75','20-25']
@interact(Age = age)
def f(Age):      
    df2= df[df['age'].str.contains(Age)].groupby('Year').sum() 
    df2['year'] = ['2018','2019']
    return sns.barplot(x="Year", y="Count", data=df2) 

The x axis is the year, the y axis is the number of participants that year and you can interactively filter via a scroll bar and see the number of participants by age range of your choosing.

I didn't know how to do that, so I tried to 'brute force it' and added a count column of 1, so that I would get the total number of participants even when the dataset got filtered. I also brute forced the years, but some age ranges don't compete in all years so It breaks and it's not working.

Age is all the various options of age range, I put them all in a list.

CodePudding user response：

• Fix the spelling issues, since you are interchanging upper and lowercase. Also none of the values in the dataframe are numeric (both issue fixed in an edit to the OP).
• Create age with age = df.Age.unique()
• Reset the index instead of using df2['year'] = ['2018','2019']

age = df.Age.unique()
@interact(Age = age)
def f(Age):      
    df2 = df[df['Age'].str.contains(Age)].groupby('Year').count().reset_index()
    return sns.barplot(x="Year", y="Count", data=df2)

• Without adding a 'Count' column
• There's no need to add a 'Count' column to the dataframe because .groupby('Year').count() will provide the count for each existing column, in which case y= can be 'Name' or 'Age', but then the ylabel would need to be changed.

age = df.Age.unique()
@interact(Age = age)
def f(Age):      
    df2 = df[df['Age'].str.contains(Age)].groupby('Year').count().reset_index()
    p = sns.barplot(x="Year", y="Age", data=df2)
    p.set_ylabel('Count')
    return p

• Use seaborn.countplot instead

age = df.Age.unique()
@interact(Age = age)
def f(Age):      
    data = df[df['Age'].str.contains(Age)]
    return sns.countplot(data=data, x='Year')