#SDC-OMICS2014-QQ Depository of scripts used to analyze a QQ log

Note: scripts are not all extentable, some of them need to be modified.

This pepline can help you find out the questions below

Q1: General information: including distributions of message count, word count, figure count and emoji count;

Q2: Who tend to be outliers among above distributions? Who starts the most conversation?

Q3: Do people have discernable chatting pattern?

Q4: The chat pattern and its association with daily events;

Q5: Do people respond to each other equally?

Q6: To be continued...

To analyze the QQ log:

Install Perl and R;
Export QQ log and put these scripts in the same directory;
Sequentially run the following scripts: generate_data.pl
transform_data.pl
integrate_words_by_day.pl
time_serial.pl
time_freq.pl
chat.pl
These scripts will generate new files in the same directory;
Open plots.R file, change the directory path in the front line, and draw figures visualizing conversation patterns;
Open tests.R file, change the directory path in the front line, and draw statistical conclustions.

Have fun