31:T463,md`My father buys Chinese National Geography every month. I read this magazine, which is similar to National Geographic and famous for its exquisite illustrations, with my father. As a six-year-old boy, I intuitively understood that dense, text-heavy articles were dull, while stories with pictures and fewer words were much easier to understand. As I grew older, I continued to see images and visual aids extremely vital to storytelling because they can convey meaning much more powerfully than words alone. Through formal training and experience in journalism and computer science, I have learned to communicate using visualizations that bring the nuance of data to life through the impact of charts and pictures. I graduated from Columbia’s dual degree program in Computer Science and Journalism in 2020. And I joined CoinDesk as data viz analyst to further explore the possibilities of data-fueled visual innovation in storytelling.

_My father’s Chinese National Geography collections_`32:T12b2,md`I participate in multiple data projects at Columbia. In one, I designed a database system in PostgreSQL using NYC open data to display crimes, monuments, eateries, hiking trails, and other sites in all public parks at New York City. I built a user interface in Python and JavaScript for visitors to access the database, which involved add, edit, delete and search functions. _I designed a NYC park database in PostgreSQL and Python with my classmate_

In another, I designed [an interactive website](https://haoshuai999.github.io/) in D3.js and scrollama.js to explore why and how global terrorism evolved over the years by analyzing 180,000 rows of data about weapons, victims, and perpetrators of terrorist attacks in the last 50 years using pandas.

In other smaller data projects, I used pandas to clean the dataset, applied linear regression or logistic regression model on the data, and made preliminary visualizations in matplotlib or R library ggplot2. In terms of data gathering, I filed a FOIL request and got the data about mobile food vendors in New York City. I visualized datasets about NBA team performances with Tableau and felonies in NYC with D3.js and built a local server showing astronomy pictures from NASA in Flask and jQuery. In my last year at Columbia, I designed a prototype for a news summarization application and built a web application in Ruby on Rails to help journalists visualize datasets.

The following are some Python code I used to analyze a dataset about initial coin offering:
\`\`\`Python import csv import numpy as np import matplotlib.pyplot as plt import statistics # The dataset is about the size of initial coin offering (ICOs) from February 2014 to July 2018 # The dataset also includes the name and close date of ICOs, which can be used for further analysis # The following script provides the median, quartiles, the mean and standard deviation of ICO size. # Next step: A company called EOS raised $4,200 million dollar through ICO on June 1, 2018 # The second largest ICO size is only about $850 million dollar up to July 2018 # How EOS raised such amount of money is worth investigating # And Syed Shoeb, a writer for Hacker Noon, wrote an article on medium about the possibility of scam for EOS's crowdfunding (https://hackernoon.com/is-eos-a-scam-eos-ico-suffers-collusion-allegations-complete-analysis-e21e68a2a5c7) # Values are specified in million dollars # Data source: Coindesk ICO Tracker (https://www.coindesk.com/ico-tracker) # Open the csv file about ICO data data_file = open("ICO Data Export.csv","r") # Convert to DictReader data_reader = csv.DictReader(data_file) # Set a list for total ICO size total_ico_size = [] # Clean the data by removing dollar sign, dash and comma for row in data_reader: total_ico_size.append(float(row[data_reader.fieldnames[2]].replace("$","").replace("-","0").replace(",",""))) # Calculate and print the quartiles, interquartile range, upper and lower bound quartiles = np.percentile(total_ico_size, [25,50,75]) iqr = quartiles[2] - quartiles[0] k = 1.5 upper_bound = quartiles[2] + k * iqr lower_bound = quartiles[0] - k * iqr print("Median = %d" % quartiles[1]) print("First quartile = %d" % quartiles[0]) print("Third quartile = %d" % quartiles[2]) print("Upper bound = %f" % upper_bound) print("Lower bound = %f" % lower_bound) print('\n') # Calculate mean and standard deviation mean = statistics.mean(total_ico_size) sd = np.std(total_ico_size) # Store and print the mean, 1, 2 and 3 standard deviation in a list called mean_measure mean_measure = [mean] print("Mean = %f" % mean) for i in range(4): lower_sd = mean - (i * sd) upper_sd = mean + (i * sd) mean_measure.append(upper_sd) mean_measure.append(lower_sd) if i > 0: print("%d standard deviation = [%f , %f]" % (i, lower_sd, upper_sd)) # Plot the histogram and exclude the outlier $4,200 million dollar to make the histogram easy to read # But the outlier is not excluded when calculating mean, median and other statistical variables histogram = plt.hist(total_ico_size, bins=100, range=(0,850), histtype='bar', color= 'blue') # Plot the quartiles, upper bound and lower bound plt.vlines(quartiles, 0, 450, colors= 'red', linestyles='dashed', linewidth= 1) plt.vlines(upper_bound, 0, 450, colors= 'red', linestyles='dashed', linewidth= 1) plt.vlines(lower_bound, 0, 450, colors= 'red', linestyles='dashed', linewidth= 1) # Plot the mean, 1, 2 and 3 standard deviation plt.vlines(mean_measure, 0, 450, colors= 'yellow', linestyles='dashed', linewidth= 1) plt.show() \`\`\` `33:[] 2f:{"title":"Why I choose Data Journalism?","cells":[{"id":0,"value":"md`# Why I choose Data Journalism?`","pinned":false,"hidden":false,"mode":"ojs"},{"id":8,"value":"$31","pinned":false,"hidden":false,"mode":"ojs"},{"id":20,"value":"md`Since I grew up reading magazines that used both artistic expression and scientific analysis to tell thrilling and fascinating stories about the world, I was deeply curious about both. In my freshman year at the Communication University of China, I was a student in the school of computer science. I took a lot of courses about computer science, such as Advanced Mathematics, Programming Methodology, Data Structure. Programming opened a new door for my future. I wrote programs to work out math problems, practice basic operations of databases and built [a simple book management system](https://github.com/haoshuai999/Library-Management).`","pinned":false,"hidden":false,"mode":"ojs"},{"id":23,"value":"md`In the spare time, my interest was developing interesting and attractive websites in the college laboratory. I was curious about displaying information creatively, like what data journalism students did in their interactive works. My good GPA in the freshman year made it possible for me to switch majors from information security to journalism, the most well-known major in the university. However, it was hard for me to get used to the differences between art and science. I was neither willing to give up my interest in journalism nor happy to stop learning programming for my future career.\n

\n_The interactive map I designed for 2017 Data Journalism Competition of China_`","pinned":false,"hidden":false,"mode":"ojs"},{"id":25,"value":"md`During the last three years of college, I explored programming, data journalism, and visualization through self-study while maintaining a superb academic record. In my sophomore year, I learned Python to scrape data from the web and Tableau to make charts in the school newspaper. I also participated in massive online open courses (MOOC) about global data journalism and created interactive web pages for a data journalism competition. When working on data projects, I was always interested in solving technical problems on the front end. I felt a strong sense of achievement every time I finished a project.\n\n

\n_I was the chief editor of the school newspaper_`","pinned":false,"hidden":false,"mode":"ojs"},{"id":27,"value":"md`In the summer of 2017, I joined Caixin Media, one of the most famous finance media companies in China, as a data journalism intern. I learned valuable knowledge about data analysis and visualization from my colleagues and published a number of news articles on topics as diverse as finance, politics, education, and science. Most notably, one article I wrote about suicide rates among older adults received over 700,000 views. The article was featured on the homepage for five days and got the attention of the company’s senior editors.`","pinned":false,"hidden":false,"mode":"ojs"},{"id":29,"value":"md`My work at the Caixin Vislab, a data journalism award-winning team, was both challenging and fulfilling. However, I also noticed some drawbacks about working in Chinese data journalism teams. Most Chinese reporters know little about programming. Some Chinese journalists are self-taught when it comes to the more technologically advanced parts of their jobs. Although I learned the basics about both programming and web design, I still want more systematic training to prepare myself for an industry where data journalism knowledge and training are scarce and sorely lacking.\n

\n_Caixin Vislab won 2018 Best Data Journalism Team in two consecutive years_`","pinned":false,"hidden":false,"mode":"ojs"},{"id":31,"value":"md`Columbia University is one of the best universities in journalism and engineering education. It is also regarded as an ideal school for data journalists around the world. I took courses about Data Science and Computational Journalism, which are vital to my technical and reporting skills.`","pinned":false,"hidden":false,"mode":"ojs"},{"id":33,"value":"$32","pinned":false,"hidden":false,"mode":"ojs"},{"id":82,"value":"md`In summer 2019, my interest in data and web design was further cultivated at USA TODAY Network, where I worked as an interactive graphics intern. I collected, cleaned, and analyzed data in Python for multiple stories and created visualizations and animations in D3.js. Collaborating with journalists, designers, and programmers, I developed a scrollable interactive webpage about the history of slavery which received more than 300,000 pageviews.

\n\n_The Twitter account for USA TODAY's graphics team_`","pinned":false,"hidden":false,"mode":"ojs"},{"id":104,"value":"md`In June 2020, I joined CoinDesk's research team. I made charts for reports, news articles and social media. I also conducted data analysis on crypto market and made handy tools on Observable to simplify the workflow of journalists. I became a full-time data journalist.`","pinned":false,"hidden":false,"mode":"ojs"},{"id":109,"value":"tweet(\"1342143954511282176\")","pinned":false,"hidden":false,"mode":"ojs"},{"id":37,"value":"md`
Nowadays, I am confident that my robust interdisciplinary academic and professional experience has fully prepared me for the unique challenges and opportunities I will encounter through my future career as a data journalist.`","pinned":false,"hidden":false,"mode":"ojs"},{"id":107,"value":"import {tweet} from \"@mbostock/tweet\"","pinned":false,"hidden":false,"mode":"ojs"}],"files":"$W33","stdlib":"1","titling":"auto"}