I'm a beginner with pandas and python in general. I would like to make a scatter plot of the column "vol" in a way that the values in 'x' axis are the ones that correspond to '1' in the column "reg", and 'y' axis are values of "vol" that correspond to '0' in "reg"
I'd appreciate your help.
| vol | dx | reg |
|---|---|---|
| 4324.208797 | CN | 1 |
| 3805.078032 | CN | 1 |
| 3820.867115 | CN | 1 |
| 3657.034962 | CN | 1 |
| 3967.540763 | CN | 1 |
| 202822.164817 | MCI | 0 |
| 240965.499488 | MCI | 0 |
| 258301.119915 | MCI | 0 |
| 220183.190232 | MCI | 0 |
| 212202.300552 | MCI | 0 |
Try the following:
import matplotlib.pyplot as plt
plt.scatter(df[df.reg==1]['vol'], df[df.reg==0]['vol'])
plt.show()
An alternative is to use pandas even for plotting as matplotlib requires too much work. But to do that, dataframe needs to be transformed into wide format for pandas plotting to work. Then plotting comes for free:
vol = df.groupby('reg').apply(lambda g: g.reset_index(drop=True)).unstack('reg')['vol']
vol.plot(kind='scatter',x=1,y=0, title='scatter plot')
groupby...reset simply provides a common index across reg series 0..4. unstack transforms it into columns/wide format that works with pandas. Here is the final dataframe that is ready for pandas plotting.
