Dataset = bird strikes on aircraft, available from Kaggle

import pandas as pd
import matplotlib.pyplot as plt

df = pd.read_csv('database.csv')

counts = df['Incident Year'].value_counts(sort=False)
counts.plot(kind='bar')
plt.show()

Passing the parameter sort=False ensures the dates on the x axis remain in the correct order, the default value is True which will result is incorrectly ordered dates.

Word clouds give only limited insight but they are popular. In python the package 'wordcloud' is popular. Below is an example of how to generate a word cloud. Once again I'm using Trump's twitter account as a source of text data.

from wordcloud import WordCloud
import matplotlib.pyplot as plt

text = '''Just leaving Florida. Big crowds of enthusiastic supporters lining the 
road that the FAKE NEWS media refuses to mention. Very dishonest!
Congratulations Stephen Miller- on representing me this morning on the 
various Sunday morning shows. Great job!
I know Mark Cuban well. He backed me big-time but I wasn't interested 
in taking all of his calls.He's not smart
 enough to run for president!
Bernie Sanders was cut off for using the term fake news to describe 
the network. They said technical difficulties!
72% of refugees admitted into U.S. (2/3 -2/11) during COURT BREAKDOWN 
are from 7 countries: SYRIA, IRAQ, SOMALIA, IRAN, SUDAN, LIBYA & YEMEN
The crackdown on illegal criminals is merely the keeping of my campaign 
promise. Gang members, drug dealers & others are being removed!
A working dinner tonight with Prime Minister Abe of Japan, and his 
representatives, at the Winter White House (Mar-a-Lago). Very good talks!
I am so proud of my daughter Ivanka. To be abused and treated so badly by 
the media, and to still hold her head so high, is truly wonderful!
'Majority in Leading EU Nations Support Trump-Style Travel Ban' 
Poll of more than 10,000 people in 10 countries
My daughter Ivanka has been treated so unfairly by Nordstrom. She is a great 
person -- always pushing me to do the right thing! Terrible!
If the U.S. does not win this case as it so obviously should, 
we can never have the security and safety to which we are entitled. Politics!
It is a disgrace that my full Cabinet is still not in place, the longest such 
delay in the history of our country. Obstruction by Democrats!
The threat from radical Islamic terrorism is very real, just look at what 
is happening in Europe and the Middle-East. Courts must act fast!
I call my own shots, largely based on an accumulation of data, and 
everyone knows it. Some FAKE NEWS media, in order to marginalize, lies!'''

word_cloud = WordCloud().generate(text)

plt.imshow(word_cloud)
plt.axis("off")

word_cloud = WordCloud(max_font_size=40).generate(text)
plt.figure()
plt.imshow(word_cloud)
plt.axis("off")
plt.show()

The above generates this word cloud:

​I am using some Trump tweets, some recent some older. Documentation for Textblob is available here, and on Github
Code used:

from textblob import TextBlob

with open('cleaned_tweets.txt') as ip_file:
    text = ip_file.readlines()
    for tweet in text:
        print('tweet: ' + tweet)
        print('\n')
        analysis = TextBlob(tweet)
        if analysis.sentiment.polarity > 0:
            print('positive')
        elif analysis.sentiment.polarity == 0:
            print('neutral')
        else:
            print('negative')
        print('polarity: ' + str(analysis.sentiment.polarity))
        print('subjectivity: ' + str(analysis.subjectivity))
        print('*******')

Results:

example 1:

tweet: My daughter Ivanka has been treated so unfairly by Nordstrom. She is a great person -- always pushing me to do the right thing! Terrible!

negative
polarity: -0.0857142857143
subjectivity: 0.821428571429

Comments: The polarity is slightly less than zero so overall it is negative. This seems reasonable as part of the tweet is very positive however the overall message is negative against Nodstrom. A subjectivity score close to 1.0 suggests a subjective message again this seems fair, the tweet is thin on facts and is mostly Trump's opinions.


example 2:

tweet: 'Majority in Leading EU Nations Support Trump-Style Travel Ban' 

neutral
polarity: 0.0
subjectivity: 0.0

Comments: Again I think this is a good result. The tweet is neutral and factual i.e. objective rather than subjective. Note that when I say factual I am not necessarily agreeing with the data Trump is quoting.


example 3:

tweet: It is a disgrace that my full Cabinet is still not in place, the longest such delay in the history of our country. Obstruction by Democrats!

positive
polarity: 0.175
subjectivity: 0.525

Comments: Here the sentiment is incorrectly marked as positive.

example 4:

tweet: I love you North Carolina- thank you for your amazing support! Get out and http://VOTE.GOP  tomorrow!

positive
polarity: 0.71875
subjectivity: 0.75

Comments: Good result.

Overall the package produced good results but sometimes it was incorrect. But it does get an A+ for ease of use.

This follows om from the script in the last post, this time instead of printing out the data I load it into a table:

from bs4 import BeautifulSoup 
import urllib               
import re
import datetime
import sqlite3 as lite

con = lite.connect('jobs.db')

firstURL = 'http://www.nijobs.com/Data-Analyst-Jobs-in-Belfast'
url_stem = 'http://www.nijobs.com'
htmlPage = urllib.urlopen(firstURL).read()
soup = BeautifulSoup(htmlPage, "lxml")
jobs = soup.find_all('div', { "class" : "job-result-title" })
for job in jobs:
    title = job.find('a',href=True)
    title_txt = title.get_text().lower()
    if re.search('data analyst', title_txt) is not None:
        with con:
            link_url = url_stem + title['href']
            ts = datetime.datetime.now()
            cur = con.cursor()
            cur.execute('INSERT INTO jobs(title, link, ts) VALUES(?,?,?)',(title_txt,link_url,ts))
​            con.commit()

The following script will scrape some data from a jobs website. Note that with web scrapping there is no one-size-fits-all solution. Each web page has a different layout, you need to study the underlying HTML to figure out where the data you need is stored. Look for patterns and identifiers such as id or class names on <div>, <a> or <p> tags. BeautifulSoup does most of the heavy lifting for you, you need urlib (for python 2) to get the HTML object. The re module enables regular expression searches.

​from bs4 import BeautifulSoup 
import urllib               
import re
import datetime
 

firstURL = 'http://www.nijobs.com/Data-Analyst-Jobs-in-Belfast'
url_stem = 'http://www.nijobs.com'
htmlPage = urllib.urlopen(firstURL).read()
soup = BeautifulSoup(htmlPage, "lxml")
jobs = soup.find_all('div', { "class" : "job-result-title" })
for job in jobs:
    title = job.find('a',href=True)
    title_txt = title.get_text().lower()
    if re.search('data analyst', title_txt) is not None:
        print(title_txt)
        print(url_stem + title['href'])
        print(datetime.datetime.now())


This script outputs data as:
​
data analyst
http://www.nijobs.com/Data-analyst-Job-1249211.aspx
2017-02-05 14:19:11.811000
data analyst
http://www.nijobs.com/Data-Analyst-Job-1251895.aspx
2017-02-05 14:19:11.839000
business data analyst
http://www.nijobs.com/Business-Data-Analyst-Job-1252814.aspx
2017-02-05 14:19:11.867000
data analyst (consumer insight)
http://www.nijobs.com/Data-Analyst-Consumer-Insight-Job-1250744.aspx
2017-02-05 14:19:11.895000
data analyst – ktp associate
http://www.nijobs.com/Data-Analyst-KTP-Associate-Job-1253389.aspx
2017-02-05 14:19:11.923000
talent data analyst
http://www.nijobs.com/Talent-Data-Analyst-Job-1251769.aspx
2017-02-05 14:19:11.950000
data analyst (contract)
http://www.nijobs.com/Data-Analyst-contract-Job-1251923.aspx
2017-02-05 14:19:11.977000

The dataset is a small set, it contains data on a variable star call AA Cas in the constellation of Cassiopeia. Some stars are variable - they do not remain at the same brightness in some cases because they are unstable. The data is available from the British Astronomical Association variable star section, here. The dates are Julian Date - a standard used in Astrophysics. Using pandas it is possible to create a simple scatter plot:

import pandas as pd

df = pd.read_csv('AA Cas_20170201_204530.csv')
df = df.drop(df.columns[[0,2,3,5,7,8,9,10,11,12,13]], axis=1)

df.plot(x='Julian_Date', y='Magnitude',kind='scatter')
plt.show()

There are some problems with this graph. First the y-axis is back-to-front. A stars apparent brightness is expressed as a measure called magnitude, this is a log scale with lower numbers being brighter. So rather than having the y axis going from about 7.5 up we want the higher numbers - dimmer stars - lower down. This can be done by adding ylim=[9.5,7.0] to the plot parameters.

​import pandas as pd

df = pd.read_csv('AA Cas_20170201_204530.csv')
df = df.drop(df.columns[[0,2,3,5,7,8,9,10,11,12,13]], axis=1)

df.plot(x='Julian_Date', y='Magnitude',kind='scatter',ylim=[9.5,7.0])
plt.show()

​The above produces the following graph:

This is better, higher points on the graph now represent brighter magnitudes for the star. There is still one problem with the graph - all the observations were made visually not electronically and they were made by different people so it would maybe be useful to use different colours to represent the different people. This can be done with the following changes to the script:

import pandas as pd
import matplotlib.pyplot as plt

df = pd.read_csv('AA Cas_20170201_204530.csv')
df = df.drop(df.columns[[0,2,3,5,7,8,9,10,11,12,13]], axis=1)

colors = {'T Markham':'red','J Toone':'blue','D Scanlan':'green','R K Hunt':'black','T L Heywood':'orange'}
fig, ax = plt.subplots()
ax.scatter(df['Julian_Date'],df['Magnitude'], c=df['Observer_Name'].apply(lambda x: colors[x]))
ax.set_ylim([9.5,7])
plt.show()

​This produces the following graph:

This can be further improved by adding a legend to explain the different colours, a title and axis labels.