Hacker News Project¶
We'll compare these two types of posts to determine the following:¶
- Do
Ask HNorShow HNreceive more comments on average? - Do posts created at a certain time receive more comments on average?
Table Headers :¶
- id : post unique identifier from Hacker News
- title : title of the post
- url : URL of the post
- num-points : number of points, upvotes minus downvotes
- num-comments : The number of comments on the post
- author : The username who submitted the post
- created-at : The date and time the post was submitted
In [1]:
import csv
import datetime as dt
In [2]:
with open('hacker_news.csv') as f:
hn_data = list(csv.reader(f))
headers = hn_data.pop(0)
print(headers)
print(hn_data[0:2])
print(len(hn_data))
In [3]:
ask_posts = [i for i in hn_data if i[1].lower().startswith('ask hn')]
show_posts = [i for i in hn_data if i[1].lower().startswith('show hn')]
othr_posts = [i for i in hn_data if not i[1].lower().startswith('show hn') and not i[1].lower().startswith('ask hn')]
print(len(ask_posts ))
print(len(show_posts))
print(len(othr_posts))
print(len(ask_posts)+len(show_posts)+len(othr_posts))
In [4]:
total_ask_comments = sum([int(i[4]) for i in ask_posts])
averg_ask_comments = total_ask_comments / len(ask_posts)
total_show_comments = sum([int(i[4]) for i in show_posts])
averg_show_comments = total_show_comments / len(show_posts)
total_othr_comments = sum([int(i[4]) for i in othr_posts])
averg_othr_comments = total_othr_comments / len(othr_posts)
print(total_ask_comments)
print(averg_ask_comments)
print(total_show_comments)
print(averg_show_comments)
print(total_othr_comments)
print(averg_othr_comments)
From the numbers above, we can see that ask and show posts recieve less comments. than other posts,
and ask posts recieve more comments than show posts
determine if ask posts created at a certain time are more likely to attract comments¶
- Calculate the amount of ask posts created in each hour of the day, along with the number of comments received.
- Calculate the average number of comments ask posts receive by hour created.
In [5]:
print(headers)
print(ask_posts[0])
In [6]:
for i in ask_posts:
i[6] = dt.datetime.strptime(i[6],'%m/%d/%Y %H:%M')
In [7]:
# creating empty dic with day hours as keys
posts_per_hour = {i:[] for i in range(24)}
print(posts_per_hour)
print(len(posts_per_hour))
In [8]:
for post in ask_posts:
posts_per_hour[post[6].hour].append(post)
In [9]:
print(len(posts_per_hour[0]))
print(len(posts_per_hour))
In [10]:
count_posts_per_hour = {i:[] for i in range(24)}
tot_comment_per_hour = {i:[] for i in range(24)}
for hour,posts in posts_per_hour.items():
count_posts_per_hour[hour] = len(posts)
tot_comment_per_hour[hour] = sum([int(post[4]) for post in posts])
print('Amount of posts for each hour:')
for i in count_posts_per_hour: print(i, ' >>> ', count_posts_per_hour[i])
print('\nTotal comments for each hour:')
for i in tot_comment_per_hour:print(i, ' >>> ', tot_comment_per_hour[i])
print('\nAvg number of comments ask posts receive by hour created:')
for i in tot_comment_per_hour:print('{:02d} >>> {:.2f}'.format(i, tot_comment_per_hour[i]/count_posts_per_hour[i]))
In [11]:
%matplotlib inline
import matplotlib.pyplot as plt
avg_by_hour = {}
for i in tot_comment_per_hour:
avg_by_hour[i] = tot_comment_per_hour[i]/count_posts_per_hour[i]
pl = plt.bar(avg_by_hour, avg_by_hour.values())
In [12]:
lol = sorted(
[ [v, k] for k,v in avg_by_hour.items() ],
reverse = True
)
In [13]:
print("Top 5 Hours for Ask Posts Comments:")
[i[1] for i in lol[:5]]
Out[13]:
In [14]:
for i in lol:
print(
dt.datetime(2020,1,1,i[1]).strftime('%H:%M') +
' : {:.2f} average comments per post'.format(i[0])
)