书写，是通往另一个世界的门。
以前跟一个朋友聊天时，她说她最羡慕的女人就是三毛：“跟她爱的人在一起，去了她想去的地方，而且还能把这些点点滴滴都用美丽的文字写下来。“
Time series visualization.
Things I learned:
Have to use numpy or datetime to transform the date column to datetime type. If use pd.to_datetime() method, somehow matplotlib won’t recognize it;
In a recent task I need to generate a 2-column date matrix, for each date in the 1st column, there will be N rows containing N previous dates. Something look like:
| date1 | date2 | |
|---|---|---|
| 2017-10-01 | 2017-10-01 | |
| 2017-10-01 | 2017-09-30 | |
| 2017-10-01 | 2017-09-29 | |
| 2017-10-02 | 2017-10-02 | |
| 2017-10-02 | 2017-10-01 | |
| 2017-10-02 | 2017-09-30 |