1 Background¶
The dataset used in this project is a Kaggle open dataset named Goodreads-books. It provides a comprehensive list of all books listed in goodreads. This dataset was last updated on March 09, 2020.
We are trying to answer the following questions:
- Which top 5 books have the highest average rating and from which publisher?
- How many books were published since 1990?
2 Data Preparation¶
There are 4 lines with unmatched column size, to successfully load using pandas, these 4 lines are skipped.
import warnings
warnings.filterwarnings('ignore')
import pandas as pd
from matplotlib import pyplot as plt
plt.style.use('ggplot')
df = pd.read_csv('C:/Users/Elle/Desktop/jupyter_folder/datasets/books.csv', error_bad_lines = False)
df.head()
2.1 The Dataset¶
The following provides a basic information on the interested dataset. Below shows that there are 11,123 entries.
df['bookID'].nunique()
df.info()
2.2 Data Cleaning¶
Notice the data type of 'publication_date' column is type object, since we are going to work with dates, it is better to convert the data type to datetime format.
df['publication_date'] = pd.to_datetime(df['publication_date'], format='%m/%d/%Y', errors='coerce')
df.info()
tempDf = df.loc[df['ratings_count'] >= 50]
hi_avg_rating = tempDf.groupby(['title','publisher'])['average_rating'].max().sort_values(ascending=False).head(5).reset_index()
hi_avg_rating
fig, ax = plt.subplots(figsize = (5,3))
ax.set_title('Top 5 books with their ratings', fontsize = 12, weight = 'bold')
ax.set_ylabel('Book Title')
ax.set_xlabel('Average ratings')
ax.barh(hi_avg_rating['title'],hi_avg_rating['average_rating'])
for i in ['top','bottom','left','right']:
ax.spines[i].set_visible(False)
ax.xaxis.set_ticks_position('none')
ax.yaxis.set_ticks_position('none')
ax.grid(False)
ax.set_facecolor("white")
ax.invert_yaxis()
for j in ax.patches:
plt.text(j.get_width()+0.2, j.get_y()+0.5,
str(round((j.get_width()),2)), fontsize = 10, color = 'black')
plt.show()
3.2 How many books were published since 1990?¶
As shown below, 10,168 unique books were published since year 1990. A 'publication_year' column is added to provide a better visualization chart.
df_pub = df.loc[df['publication_date'] >= '01/01/1990']
df_pub['publication_year'] = pd.DatetimeIndex(df_pub['publication_date']).year
df_pub.shape
df_pub_temp = df_pub.groupby(['publication_year'])['bookID'].count().reset_index()
df_pub_temp
df_pub_temp.describe()
fig_pub, ax_pub = plt.subplots()
fig_pub = plt.figure(figsize = (5,3))
ax_pub.set_title('Number of published books since 1990 ', fontsize = 12, weight = 'bold')
ax_pub.set_xlabel('Publication Year')
ax_pub.set_ylabel('Number of Books')
ax_pub.grid(False)
ax_pub.set_facecolor("white")
ax_pub.bar(df_pub_temp['publication_year'],df_pub_temp['bookID'])
plt.show()
One interesting observation from the above bar chart is that the rate at which the books were published was at an increasing slope between 1990 to 2006. At around 2007, the slope started to decline.