Introduction

This project focuses on analyzing user data from Bumble, a well-known dating app, to uncover valuable insights about its users. By exploring the information users share in their profiles such as demographics, lifestyle habits, and preferences we aim to understand their behavior and trends better. These insights will help Bumble’s product and marketing teams make informed decisions to improve user engagement, fine-tune the matchmaking algorithms, and offer personalized features. Ultimately, the goal is to create a more satisfying experience for Bumble users and support the platform's growth.

Installing Pandas

%pip install pandas

Requirement already satisfied: pandas in /usr/local/lib/python3.10/dist-packages (2.2.2)
Requirement already satisfied: numpy>=1.22.4 in /usr/local/lib/python3.10/dist-packages (from pandas) (1.26.4)
Requirement already satisfied: python-dateutil>=2.8.2 in /usr/local/lib/python3.10/dist-packages (from pandas) (2.8.2)
Requirement already satisfied: pytz>=2020.1 in /usr/local/lib/python3.10/dist-packages (from pandas) (2024.2)
Requirement already satisfied: tzdata>=2022.7 in /usr/local/lib/python3.10/dist-packages (from pandas) (2024.2)
Requirement already satisfied: six>=1.5 in /usr/local/lib/python3.10/dist-packages (from python-dateutil>=2.8.2->pandas) (1.17.0)

As the next step imported Pandas as pd , numpy as np , matplotlib.pyplot as plt , seaborn as sns. Pyplot and Seaborn are useful to create visualizations further.

import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
import seaborn as sns

from google.colab import drive
drive.mount('/content/drive')

Drive already mounted at /content/drive; to attempt to forcibly remount, call drive.mount("/content/drive", force_remount=True).

import pandas as pd

# Provide the full path to your dataset
dataset_path = '/content/drive/MyDrive/Grad Project Dataset/bumble (1).csv'
df = pd.read_csv(dataset_path)

# Display the first few rows of the dataframe
print(df.head())

   age     status gender       body_type               diet    drinks  \
0   22     single      m  a little extra  strictly anything  socially   
1   35     single      m         average       mostly other     often   
2   38  available      m            thin           anything  socially   
3   23     single      m            thin         vegetarian  socially   
4   29     single      m        athletic                NaN  socially   

                           education            ethnicity  height  income  \
0      working on college/university         asian, white    75.0      -1   
1              working on space camp                white    70.0   80000   
2     graduated from masters program                  NaN    68.0      -1   
3      working on college/university                white    71.0   20000   
4  graduated from college/university  asian, black, other    66.0      -1   

                           job       last_online  \
0               transportation  2012-06-28-20-30   
1         hospitality / travel  2012-06-29-21-41   
2                          NaN  2012-06-27-09-10   
3                      student  2012-06-28-14-22   
4  artistic / musical / writer  2012-06-27-21-26   

                          location                       pets  \
0  south san francisco, california  likes dogs and likes cats   
1              oakland, california  likes dogs and likes cats   
2        san francisco, california                   has cats   
3             berkeley, california                 likes cats   
4        san francisco, california  likes dogs and likes cats   

                                   religion  \
0     agnosticism and very serious about it   
1  agnosticism but not too serious about it   
2                                       NaN   
3                                       NaN   
4                                       NaN   

                                 sign  \
0                              gemini   
1                              cancer   
2  pisces but it doesn’t matter   
3                              pisces   
4                            aquarius   

                                              speaks  
0                                            english  
1  english (fluently), spanish (poorly), french (...  
2                               english, french, c++  
3                           english, german (poorly)  
4                                            english  

We used Drop function here to remove the duplicates from the dataset.

df= df.drop_duplicates()
df

	age	status	gender	body_type	diet	drinks	education	ethnicity	height	income	job	last_online	location	pets	religion	sign	speaks
0	22	single	m	a little extra	strictly anything	socially	working on college/university	asian, white	75.0	-1	transportation	2012-06-28-20-30	south san francisco, california	likes dogs and likes cats	agnosticism and very serious about it	gemini	english
1	35	single	m	average	mostly other	often	working on space camp	white	70.0	80000	hospitality / travel	2012-06-29-21-41	oakland, california	likes dogs and likes cats	agnosticism but not too serious about it	cancer	english (fluently), spanish (poorly), french (...
2	38	available	m	thin	anything	socially	graduated from masters program	NaN	68.0	-1	NaN	2012-06-27-09-10	san francisco, california	has cats	NaN	pisces but it doesn’t matter	english, french, c++
3	23	single	m	thin	vegetarian	socially	working on college/university	white	71.0	20000	student	2012-06-28-14-22	berkeley, california	likes cats	NaN	pisces	english, german (poorly)
4	29	single	m	athletic	NaN	socially	graduated from college/university	asian, black, other	66.0	-1	artistic / musical / writer	2012-06-27-21-26	san francisco, california	likes dogs and likes cats	NaN	aquarius	english
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
59941	59	single	f	NaN	NaN	socially	graduated from college/university	NaN	62.0	-1	sales / marketing / biz dev	2012-06-12-21-47	oakland, california	has dogs	catholicism but not too serious about it	cancer and it’s fun to think about	english
59942	24	single	m	fit	mostly anything	often	working on college/university	white, other	72.0	-1	entertainment / media	2012-06-29-11-01	san francisco, california	likes dogs and likes cats	agnosticism	leo but it doesn’t matter	english (fluently)
59943	42	single	m	average	mostly anything	not at all	graduated from masters program	asian	71.0	100000	construction / craftsmanship	2012-06-27-23-37	south san francisco, california	NaN	christianity but not too serious about it	sagittarius but it doesn’t matter	english (fluently)
59944	27	single	m	athletic	mostly anything	socially	working on college/university	asian, black	73.0	-1	medicine / health	2012-06-23-13-01	san francisco, california	likes dogs and likes cats	agnosticism but not too serious about it	leo and it’s fun to think about	english (fluently), spanish (poorly), chinese ...
59945	39	single	m	average	NaN	socially	graduated from masters program	white	68.0	-1	medicine / health	2012-06-29-00-42	san francisco, california	likes dogs and likes cats	catholicism and laughing about it	gemini and it’s fun to think about	english

59946 rows × 17 columns

Part 1: Data Cleaning

missing_values = df.isnull().sum()
missing_values

	0
age	0
status	0
gender	0
body_type	5296
diet	24395
drinks	2985
education	6628
ethnicity	5680
height	3
income	0
job	8198
last_online	0
location	0
pets	19921
religion	20226
sign	11056
speaks	50

dtype: int64

# Get the index of null values in the 'height' column
null_indexes = df[df['height'].isnull()].index
print(null_indexes)

Index([36428, 54002, 58983], dtype='int64')

Question 2 : Are there columns where more than 50% of the data is missing? Drop those columns where missing values are >50%.

percentage_of_missing_values = ((missing_values)/ len(df)) * 100
percentage_of_missing_values

	0
age	0.000000
status	0.000000
gender	0.000000
body_type	8.834618
diet	40.694959
drinks	4.979482
education	11.056618
ethnicity	9.475194
height	0.005005
income	0.000000
job	13.675641
last_online	0.000000
location	0.000000
pets	33.231575
religion	33.740366
sign	18.443266
speaks	0.083408

dtype: float64

missing_data = pd.DataFrame({ 'Missing Values': missing_values,'Percentage Missing': percentage_of_missing_values})
missing_data

	Missing Values	Percentage Missing
age	0	0.000000
status	0	0.000000
gender	0	0.000000
body_type	5296	8.834618
diet	24395	40.694959
drinks	2985	4.979482
education	6628	11.056618
ethnicity	5680	9.475194
height	3	0.005005
income	0	0.000000
job	8198	13.675641
last_online	0	0.000000
location	0	0.000000
pets	19921	33.231575
religion	20226	33.740366
sign	11056	18.443266
speaks	50	0.083408

Here we created a dataframe to get a clear understanding of column wise missing values and missing percentage. As there are no columns with morethan 50% of missing data, we don't need to drop any columns.

Question 3 : Missing numerical data (e.g., height, income) should be handled by imputing the median value of height and income for the corresponding category, such as gender, age group, or location. This ensures that the imputed values are contextually relevant and reduce potential biases in the analysis.

calculate_median = df.groupby(["gender"])["height"].transform("median")
df["height"] = df["height"].fillna(calculate_median)
print(calculate_median)

0        70.0
1        70.0
2        70.0
3        70.0
4        70.0
         ... 
59941    65.0
59942    70.0
59943    70.0
59944    70.0
59945    70.0
Name: height, Length: 59946, dtype: float64

(II) Data Types

Question 1 : Are there any inconsistencies in the data types across columns (e.g., numerical data stored as strings)?

df.dtypes

	0
age	int64
status	object
gender	object
body_type	object
diet	object
drinks	object
education	object
ethnicity	object
height	float64
income	int64
job	object
last_online	object
location	object
pets	object
religion	object
sign	object
speaks	object

dtype: object

Question 2 : Which columns require conversion to numerical data types for proper analysis (e.g., income)?

df["last_online"]= pd.to_datetime(df["last_online"] , format= "%Y-%m-%d-%H-%M")
df["last_online"]

	last_online
0	2012-06-28 20:30:00
1	2012-06-29 21:41:00
2	2012-06-27 09:10:00
3	2012-06-28 14:22:00
4	2012-06-27 21:26:00
...	...
59941	2012-06-12 21:47:00
59942	2012-06-29 11:01:00
59943	2012-06-27 23:37:00
59944	2012-06-23 13:01:00
59945	2012-06-29 00:42:00

59946 rows × 1 columns

dtype: datetime64[ns]

df.dtypes

	0
age	int64
status	object
gender	object
body_type	object
diet	object
drinks	object
education	object
ethnicity	object
height	float64
income	int64
job	object
last_online	datetime64[ns]
location	object
pets	object
religion	object
sign	object
speaks	object

dtype: object

III) Outliers

Question 1 : Are there any apparent outliers in numerical columns such as age, height, or income? What are the ranges of values in these columns?

Used Interquartile Range (IQR) to calculate the outliers in numerical columns. Calculated lower_bound and upper_bound using formulas, then calulated the outliers. For range we found out max and min values and then substracted max and min.

column_names = ["income", "age", "height"]
for columns in column_names:
     Q1 = df[columns].quantile(0.25)
     Q3 = df[columns].quantile(0.75)
     iqr = Q3-Q1
     lower_bound = Q1-1.5*iqr
     upper_bound = Q3+1.5*iqr
     outlier = (df[columns] <= lower_bound) | (df[columns] >= upper_bound)
     outlier_values = df[outlier]
     range_value = df[columns].max() - df[columns].min()
     print(f"\n Outliers in {columns}")
     print(f"Range of {columns}: Min = {df[columns].min()}, Max = {df[columns].max()}, Range = {range_value}")
     if columns == "age":
          outlier_values = outlier_values.sort_values(by="age", ascending=True)
     elif columns == "income":
          outlier_values = outlier_values.sort_values(by="income", ascending=False)
     print(outlier_values[columns])

 Outliers in income
Range of income: Min = -1, Max = 1000000, Range = 1000001
37173    1000000
37884    1000000
39222    1000000
55335    1000000
57466    1000000
          ...   
22164         -1
22165         -1
22166         -1
22167         -1
59945         -1
Name: income, Length: 59946, dtype: int64

 Outliers in age
Range of age: Min = 18, Max = 110, Range = 92
40999     54
5394      54
14705     54
52394     54
8675      54
        ... 
52644     69
28700     69
6448      69
25324    109
2512     110
Name: age, Length: 2638, dtype: int64

 Outliers in height
Range of height: Min = 1.0, Max = 95.0, Range = 94.0
189      79.0
243      80.0
280      80.0
402      91.0
433      79.0
         ... 
58147    94.0
58286    84.0
59038    79.0
59067    79.0
59697    79.0
Name: height, Length: 285, dtype: float64

Question 2 : Any -1 values in numerical columns like income should be replaced with 0, as they may represent missing or invalid data. bold text

columns = df[["income", "age", "height"]]
columns.replace(-1, 0, inplace= True)
print(columns)

       income  age  height
0           0   22    75.0
1       80000   35    70.0
2           0   38    68.0
3       20000   23    71.0
4           0   29    66.0
...       ...  ...     ...
59941       0   59    62.0
59942       0   24    72.0
59943  100000   42    71.0
59944       0   27    73.0
59945       0   39    68.0

[59946 rows x 3 columns]

<ipython-input-16-fa4e58afb01b>:2: SettingWithCopyWarning: 
A value is trying to be set on a copy of a slice from a DataFrame

See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy
  columns.replace(-1, 0, inplace= True)

Question 3 : For other outliers, rather than deleting them, calculate the mean and median values using only the middle 80% of the data (removing extreme high and low values). This approach ensures that outliers do not disproportionately impact the analysis while retaining as much meaningful data as possible.

column_names = ["income", "age", "height"]
for columns in column_names:
    Q1 = df[columns].quantile(0.10)
    Q3 = df[columns].quantile(0.90)
    filtered_values = df[(df[columns] > Q1) & (df[columns] < Q3)]
    mean_values = filtered_values[columns].mean()
    median_values = filtered_values[columns].median()
    print(f"\n {columns}: filtered_median: {median_values}, filtered_mean: {mean_values}")

 income: filtered_median: 20000.0, filtered_mean: 26109.89010989011

 age: filtered_median: 30.0, filtered_mean: 31.357685563997663

 height: filtered_median: 68.0, filtered_mean: 68.25442646465552

IV) Missing Data Visualization

Question 1 : Create a heatmap to visualize missing values across the dataset. Which columns show consistent missing data patterns?

missing_values = df.isnull().sum()
print(missing_values)
plt.figure(figsize=(12, 8))
sns.heatmap(df.isnull(), cbar = False, cmap = "inferno")
plt.title('Heatmap of Missing Data')

age                0
status             0
gender             0
body_type       5296
diet           24395
drinks          2985
education       6628
ethnicity       5680
height             0
income             0
job             8198
last_online        0
location           0
pets           19921
religion       20226
sign           11056
speaks            50
dtype: int64

Text(0.5, 1.0, 'Heatmap of Missing Data')

Part 2: Data Processing

(I) Binning and Grouping : Question 1 : Bin the age column into categories such as "18-25", "26-35", "36-45", and "46+" to create a new column, age_group. How does the distribution of users vary across these age ranges?

bins = [18,25, 35, 45, 110]
label = ["18-25", "26-35", "36-45","46+"]
df["age_group"] = pd.cut(df["age"], bins, labels= label)
df["age_group"]
sns.countplot(x='age_group', data=df, palette="mako")
plt.title("Distribution of Users Across Age Groups")

<ipython-input-19-bf52aafc422b>:5: FutureWarning: 

Passing `palette` without assigning `hue` is deprecated and will be removed in v0.14.0. Assign the `x` variable to `hue` and set `legend=False` for the same effect.

  sns.countplot(x='age_group', data=df, palette="mako")

Text(0.5, 1.0, 'Distribution of Users Across Age Groups')

Question 2 : Group income into categories like "Low Income," "Medium Income," and "High Income" based on meaningful thresholds (e.g., quartiles). What insights can be derived from these groups?

# Group income into categories like "Low Income," "Medium Income," and "High Income" based on meaningful thresholds (e.g., quartiles).
# What insights can be derived from these groups?
Q1 = df["income"].quantile(0.25)
Q2 = df["income"].quantile(0.50)
Q3 = df["income"].quantile(0.90)
df["Income_Category"] = df["income"].apply(lambda x:"Low Income" if x <= Q1 else ("Medium Income" if Q1 < x < Q3 else "High Income"))
print(df[["income", "Income_Category"]].head(10))

   income Income_Category
0      -1      Low Income
1   80000     High Income
2      -1      Low Income
3   20000   Medium Income
4      -1      Low Income
5      -1      Low Income
6      -1      Low Income
7      -1      Low Income
8      -1      Low Income
9      -1      Low Income

df

	age	status	gender	body_type	diet	drinks	education	ethnicity	height	income	job	last_online	location	pets	religion	sign	speaks	age_group	Income_Category
0	22	single	m	a little extra	strictly anything	socially	working on college/university	asian, white	75.0	-1	transportation	2012-06-28 20:30:00	south san francisco, california	likes dogs and likes cats	agnosticism and very serious about it	gemini	english	18-25	Low Income
1	35	single	m	average	mostly other	often	working on space camp	white	70.0	80000	hospitality / travel	2012-06-29 21:41:00	oakland, california	likes dogs and likes cats	agnosticism but not too serious about it	cancer	english (fluently), spanish (poorly), french (...	26-35	High Income
2	38	available	m	thin	anything	socially	graduated from masters program	NaN	68.0	-1	NaN	2012-06-27 09:10:00	san francisco, california	has cats	NaN	pisces but it doesn’t matter	english, french, c++	36-45	Low Income
3	23	single	m	thin	vegetarian	socially	working on college/university	white	71.0	20000	student	2012-06-28 14:22:00	berkeley, california	likes cats	NaN	pisces	english, german (poorly)	18-25	Medium Income
4	29	single	m	athletic	NaN	socially	graduated from college/university	asian, black, other	66.0	-1	artistic / musical / writer	2012-06-27 21:26:00	san francisco, california	likes dogs and likes cats	NaN	aquarius	english	26-35	Low Income
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
59941	59	single	f	NaN	NaN	socially	graduated from college/university	NaN	62.0	-1	sales / marketing / biz dev	2012-06-12 21:47:00	oakland, california	has dogs	catholicism but not too serious about it	cancer and it’s fun to think about	english	46+	Low Income
59942	24	single	m	fit	mostly anything	often	working on college/university	white, other	72.0	-1	entertainment / media	2012-06-29 11:01:00	san francisco, california	likes dogs and likes cats	agnosticism	leo but it doesn’t matter	english (fluently)	18-25	Low Income
59943	42	single	m	average	mostly anything	not at all	graduated from masters program	asian	71.0	100000	construction / craftsmanship	2012-06-27 23:37:00	south san francisco, california	NaN	christianity but not too serious about it	sagittarius but it doesn’t matter	english (fluently)	36-45	High Income
59944	27	single	m	athletic	mostly anything	socially	working on college/university	asian, black	73.0	-1	medicine / health	2012-06-23 13:01:00	san francisco, california	likes dogs and likes cats	agnosticism but not too serious about it	leo and it’s fun to think about	english (fluently), spanish (poorly), chinese ...	26-35	Low Income
59945	39	single	m	average	NaN	socially	graduated from masters program	white	68.0	-1	medicine / health	2012-06-29 00:42:00	san francisco, california	likes dogs and likes cats	catholicism and laughing about it	gemini and it’s fun to think about	english	36-45	Low Income

59946 rows × 19 columns

(II) Derived Features

Question 1 : Create a new feature, profile_completeness, by calculating the percentage of non-missing values for each user profile. How complete are most user profiles, and how does completeness vary across demographics?

df["profile_completeness"] = df.notnull().sum(axis=1) / df.shape[1] * 100
print(df["profile_completeness"])
print(df.groupby("status")["profile_completeness"].mean())

0        100.000000
1        100.000000
2         84.210526
3         94.736842
4         89.473684
            ...    
59941     84.210526
59942    100.000000
59943     94.736842
59944    100.000000
59945     94.736842
Name: profile_completeness, Length: 59946, dtype: float64
status
available         91.364470
married           90.747029
seeing someone    91.115871
single            90.774803
unknown           84.210526
Name: profile_completeness, dtype: float64

(III) Unit Conversion

Question 1 : Convert the height column from inches to centimeters using the conversion factor (1 inch = 2.54 cm). Store the converted values in a new column, height_cm.

df["height_cm"] = df["height"] * 2.54
print(df[["height" , "height_cm"]])

       height  height_cm
0        75.0     190.50
1        70.0     177.80
2        68.0     172.72
3        71.0     180.34
4        66.0     167.64
...       ...        ...
59941    62.0     157.48
59942    72.0     182.88
59943    71.0     180.34
59944    73.0     185.42
59945    68.0     172.72

[59946 rows x 2 columns]

Part 3: Data Analysis

(I) Demographic Analysis Question 1 : What is the gender distribution (gender) across the platform? Are there any significant imbalances?

gender_distribution = df.groupby("gender")["gender"].count().reset_index(name = "count")
print(gender_distribution)
gender_distribution.plot.bar(x="gender", color = "lightblue")
plt.xticks(rotation = 0)
plt.title("Distribution of gender across the platform")

  gender  count
0      f  24117
1      m  35829

Text(0.5, 1.0, 'Distribution of gender across the platform')

Question 2 : What are the proportions of users in different status categories (e.g., single, married, seeing someone)? What does this suggest about the platform’s target audience?

status_proportions = df.groupby("status")["status"].count()/len(df["status"]) * 100
status_proportions_df = status_proportions.reset_index(name =  "Proportion in %")
status_proportions_df = status_proportions_df.sort_values(by = "Proportion in %", ascending = False)
print(status_proportions_df)
status_proportions.plot.bar(x="status", color = "blue")
plt.ylabel("Proportion in %")
plt.xticks(rotation = 0)
plt.title("Distribution of status across the platform")

           status  Proportion in %
3          single        92.911954
2  seeing someone         3.443099
0       available         3.111133
1         married         0.517132
4         unknown         0.016682

Text(0.5, 1.0, 'Distribution of status across the platform')

Question 3 : How does status vary by gender? For example, what proportion of men and women identify as single?

status_by_gender = df.groupby(["status", "gender"])["status"].count()/len(df["status"]) * 100
status_by_gender_df = status_by_gender.reset_index(name =  "Proportion in %")
status_by_gender_df = status_by_gender_df.sort_values(by = "Proportion in %", ascending = False)
print(status_by_gender_df)
plt.figure(figsize=(10, 6))
sns.relplot(status_by_gender_df, x= "status", y = "Proportion in %", col = "gender", kind = "scatter")

           status gender  Proportion in %
7          single      m        55.680112
6          single      f        37.231842
1       available      m         2.016815
5  seeing someone      m         1.769926
4  seeing someone      f         1.673173
0       available      f         1.094318
3         married      m         0.291929
2         married      f         0.225203
9         unknown      m         0.010009
8         unknown      f         0.006673

<seaborn.axisgrid.FacetGrid at 0x783e81f64640>

<Figure size 1000x600 with 0 Axes>

(II) Correlation Analysis

Question 1 : What are the correlations between numerical columns such as age, income, gender Are there any strong positive or negative relationships?

correlation_columns = df[["age", "income", "height"]]
correlation_data = correlation_columns.corr()
print(correlation_data)
plt.figure(figsize=(10, 6))
sns.heatmap(correlation_data, annot=True, fmt=".2f", cmap="Blues", cbar=True)
plt.title("Correlation Heatmap")
plt.show()

             age    income    height
age     1.000000 -0.001004 -0.022253
income -0.001004  1.000000  0.065048
height -0.022253  0.065048  1.000000

Question 2 : How does age correlate with income? Are older users more likely to report higher income levels?

correlation = df["age"].corr(df["income"])
print(correlation)

-0.0010038398754361403

(III) Diet and Lifestyle Analysis

Question 1 : How do dietary preferences (diet) distribute across the platform? For example, what percentage of users identify as vegetarian, vegan, or follow "anything" diets?

dietary_preferences =  df.groupby("diet")["diet"].count() / len(df["diet"]) * 100
print(dietary_preferences)

diet
anything               10.314283
halal                   0.018350
kosher                  0.018350
mostly anything        27.666567
mostly halal            0.080072
mostly kosher           0.143462
mostly other            1.679845
mostly vegan            0.563841
mostly vegetarian       5.745171
other                   0.552164
strictly anything       8.529343
strictly halal          0.030027
strictly kosher         0.030027
strictly other          0.754012
strictly vegan          0.380342
strictly vegetarian     1.459647
vegan                   0.226871
vegetarian              1.112668
Name: diet, dtype: float64

Question 2 : How do drinking habits (drinks) vary across different diet categories? Are users with stricter diets (e.g., vegan) less likely to drink?

drinking_habits = df.groupby("diet")["drinks"].count()/len(df["drinks"]) * 100
drinking_habits_df = drinking_habits.reset_index(name =  "drinks in %")
drinking_habits_df = drinking_habits_df.sort_values(by = "drinks in %", ascending = False)
print(drinking_habits_df)

                   diet  drinks in %
3       mostly anything    27.086044
0              anything     9.995663
10    strictly anything     8.319154
8     mostly vegetarian     5.539986
6          mostly other     1.626464
15  strictly vegetarian     1.416275
17           vegetarian     1.037601
13       strictly other     0.720649
7          mostly vegan     0.537150
9                 other     0.520468
14       strictly vegan     0.365329
16                vegan     0.211857
5         mostly kosher     0.140126
4          mostly halal     0.071731
11       strictly halal     0.030027
12      strictly kosher     0.030027
2                kosher     0.018350
1                 halal     0.015014

(IV) Geographical Insights

Question 1 : Extract city and state information from the location column. What are the top 5 cities and states with the highest number of users?

df[["city", "state"]] = df["location"].str.split(",", expand = True, n= 1)
df["city"] = df["city"].str.strip()
df["state"] = df["state"].str.strip()
top_cities = df["city"].value_counts(ascending = False).head(5)
top_states = df["state"].value_counts(ascending = False).head(5)
print(f"Top 5 cities with highest number of users\n {top_cities}" )
print(f"Top 5 states with highest number of users\n {top_states}")

Top 5 cities with highest number of users
 city
san francisco    31064
oakland           7214
berkeley          4212
san mateo         1331
palo alto         1064
Name: count, dtype: int64
Top 5 states with highest number of users
 state
california       59855
new york            17
illinois             8
massachusetts        5
texas                4
Name: count, dtype: int64

Question 2 : How does age vary across the top cities? Are certain cities dominated by younger or older users?

top_cities = df["city"].value_counts(ascending = False).head(5).index
top_cities_df = df[df["city"].isin(top_cities)]
age_cities = top_cities_df.groupby("city")["age"].mean().sort_values(ascending=False)
print(age_cities)

city
san mateo        33.437265
oakland          33.178819
palo alto        31.980263
san francisco    31.614312
berkeley         31.391738
Name: age, dtype: float64

Question 3 : What are the average income levels in the top states or cities? Are there regional patterns in reported income?

top_states = df["state"].value_counts(ascending = False).head(5).index
top_states_df = df[df["state"].isin(top_states)]
income_states = top_states_df.groupby("state")["income"].mean().sort_values(ascending=False)
print(income_states)

state
new york         31763.823529
california       20043.465609
massachusetts     5999.200000
texas             4999.250000
illinois            -1.000000
Name: income, dtype: float64

(V) Height Analysis

Question 1 : What is the average height of users across different gender categories?

height_gender = df.groupby("gender")["height"].mean()
print(height_gender)

gender
f    65.103869
m    70.443468
Name: height, dtype: float64

Question 2 : How does height vary by age_group? Are there noticeable trends among younger vs. older users?

height_age = df.groupby("age_group")["height"].mean()
print(height_age)

age_group
18-25    68.224532
26-35    68.406764
36-45    68.325095
46+      67.941167
Name: height, dtype: float64

<ipython-input-35-12c16651a4a6>:1: FutureWarning: The default of observed=False is deprecated and will be changed to True in a future version of pandas. Pass observed=False to retain current behavior or observed=True to adopt the future default and silence this warning.
  height_age = df.groupby("age_group")["height"].mean()

(VI) Income Analysis

Question 1 : What is the distribution of income across the platform? Are there specific income brackets that dominate? (don't count 0)

income_excluding_zero = df[df["income"] > 0]
income_excluding_zero["income"].value_counts().sort_values(ascending = False)
plt.figure(figsize=(12,6))
sns.countplot(income_excluding_zero , x = "income")

<Axes: xlabel='income', ylabel='count'>

Part 4 : Data Visualization

(I) Age Distribution Question 1 : Plot a histogram of age with a vertical line indicating the mean age. What does the distribution reveal about the most common age group on the platform?

plt.hist(df["age"], color = "Grey", edgecolor= "black")
mean_age = df["age"].mean()
plt.axvline(mean_age, color= "red", label= "mean age")
plt.xlabel("age_group")
plt.title("Distribution of age across the platform")
plt.show()

Question 2 : How does the age distribution differ by gender? Are there age groups where one gender is more prevalent?

plt.figure(figsize=(10, 4))
sns.histplot(data=df, x="age", hue="gender", palette = "mako")
plt.title("Age distribution differ by gender")

Text(0.5, 1.0, 'Age distribution differ by gender')

(II) Income and Age Question 1 : Use a scatterplot to visualize the relationship between income and age, with a trend line indicating overall patterns. Are older users more likely to report higher incomes?

sns.regplot(data= df[df["income"] > 0], x="age", y="income", scatter=True, line_kws={"color": "red"})
plt.ticklabel_format(style='plain')
plt.xlabel("Age")
plt.ylabel("Income")
plt.yticks(np.arange(20000,1000000,step=100000))
plt.title("Scatterplot of Income vs. Age with Trend Line")

Text(0.5, 1.0, 'Scatterplot of Income vs. Age with Trend Line')

Question 2 : Create boxplots of income grouped by age_group. Which age group reports the highest median income?

data_to_plot = df[df["income"] >0]
a=sns.boxplot(data_to_plot, x="age_group", y="income")
a.set(ylim=(20000,1000000))
y_ticks = np.arange(20000, 1000000, 200000)  # Start, stop, step
plt.yticks(y_ticks)
plt.title("Income vs age_group")

Text(0.5, 1.0, 'Income vs age_group')

Question 3 : Analyze income levels within gender and status categories. For example, are single men more likely to report higher incomes than single women?

income_gender_status = df.groupby(["gender", "status"])["income"].mean().reset_index()
sns.barplot(data=income_gender_status, x="status", y="income", hue="gender")
plt.title("Income levels with gender and status categories.")

Text(0.5, 1.0, 'Income levels with gender and status categories.')

(III) Pets and Preferences

Question 1 : Create a bar chart showing the distribution of pets categories (e.g., likes dogs, likes cats). Which preferences are most common?

plt.figure(figsize=(12,6))
sns.countplot(data= df , x = "pets")
plt.xticks(rotation = 90)
plt.title("Distribution of pet categories.")

Text(0.5, 1.0, 'Distribution of pet categories.')

Question 2 : How do pets preferences vary across gender and age_group? Are younger users more likely to report liking pets compared to older users?

pets_gender_age = df.groupby(["gender", "age_group", "pets"]).size().reset_index(name="count")
sns.relplot(data=pets_gender_age, x="age_group", y="count", hue="pets", col="gender", kind="scatter", palette = "Paired")
plt.figure(figsize=(12, 6))

<ipython-input-43-7a4db60f4f76>:1: FutureWarning: The default of observed=False is deprecated and will be changed to True in a future version of pandas. Pass observed=False to retain current behavior or observed=True to adopt the future default and silence this warning.
  pets_gender_age = df.groupby(["gender", "age_group", "pets"]).size().reset_index(name="count")

<Figure size 1200x600 with 0 Axes>

<Figure size 1200x600 with 0 Axes>

(IV) Signs and Personality

Question 1 : Create a pie chart showing the distribution of zodiac signs (sign) across the platform. Which signs are most and least represented? Is this the right chart? If not, replace with right chart.

zodiac_counts = df["sign"].value_counts()
zodiac_counts.plot(kind='pie')
plt.title("Distribution of Zodiac Signs")

Text(0.5, 1.0, 'Distribution of Zodiac Signs')

plt.figure(figsize=(12,6))
sns.countplot(data= df , x = "sign")
plt.xticks(rotation = 90)
plt.title("Distribution of Zodiac Signs")

Text(0.5, 1.0, 'Distribution of Zodiac Signs')

Question 2 : How does sign vary across gender and status? Are there noticeable patterns or imbalances?

gender_status = df.groupby(["gender", "status", "sign"])["status"].count().reset_index(name="count")
stacked_data = gender_status.pivot_table(index="sign", columns=["gender", "status"], values="count", aggfunc="sum")
stacked_data.plot(kind="bar", stacked=True, figsize=(20, 8), colormap="Paired")
plt.title("Distribution of Signs by Gender and Status")
plt.ylabel("Count")
plt.xlabel("Sign")
plt.xticks(rotation=90)
plt.legend(bbox_to_anchor=(1.05, 1.0), loc='upper left')

<matplotlib.legend.Legend at 0x783e79356020>

Summary of Analysis:

The analysis of Bumble user data revealed several key insights. There are significant missing values in columns such as diet and religion, while age and income data are complete. The largest group of users falls within the 26-35 age range, with more male users than female users. Profile completeness is higher for users with available data, reaching 91%, whereas the "unknown" category has lower completeness at 84%. Among the diet categories, "mostly anything" has the highest percentage of users with drinking habit at 27%. In top cities like San Mateo, the average age of users is around 33, while in New York, the average income is 31,763, which is the highest across states. Males are generally taller than females, with an average height of 70.44 inches compared to 65.10 inches. The income of most users ranges between 20,000 and 100,000, with the highest income found in the 36-45 age group ($27,786) and the lowest in the 26-35 female group (10,535). Users are mostly in their 30s, and males dominate this group. Users who like dogs and cats are more common, especially among male users. "Gemini" is the most common zodiac sign among both male and female singles.

Recommendations:

Based on these findings, Bumble could focus more on engaging the 26-35 age group, particularly male users, since they are the most active and have higher incomes. The platform should also consider adding more features related to diet preferences, as this could increase user interaction, especially for those in the "mostly anything" category. Fixing the missing data like diet and religion columns would help make recommendations more accurate. Since a lot of users like pets, especially dogs and cats, including pet preferences in the matchmaking algorithm could improve the experience. Also, since zodiac signs are popular, Bumble could use this information in marketing campaigns to engage more users.