Use of timestamp or datetime object for colorcoding of a scatterplot?

Question

I do have a list of data:

a = [[Timestamp('2015-01-01 15:00:00', tz=None), 53.0, 958.0],
     [Timestamp('2015-01-01 16:00:00', tz=None), 0.0, 900.0],
     [Timestamp('2015-01-02 11:00:00', tz=None), 543.0, 820.0], .....]

My goal is to plot the second element of each list entry vs. the third element of each list entry colorcoded by the timestamp.

import matplotlib.pyplot as plt
fig, ax = plt.subplots()
smap = ax.scatter(a[:,1], a[:,2]) 
plt.show()

As soon as I change the line for plotting to smap = ax.scatter(a[:,1], a[:,2], c = a[:,0])

I receive the error message:

'Timestamp' object has no attribute 'view'.

I think my general question is: Is there any solution in Python to plot two columns of data color-coded by date using a third column which is a timestamp or datetime-object?

Nickil Maveli · Accepted Answer · 2016-09-14 13:14:21Z

a = [[pd.Timestamp('2015-01-01 15:00:00', tz=None), 53.0, 958.0], 
     [pd.Timestamp('2015-01-01 16:00:00', tz=None), 0.0, 900.0], 
     [pd.Timestamp('2015-01-02 11:00:00', tz=None), 543.0, 820.0]]

df = pd.DataFrame(a).add_prefix('Col_')
df

df.dtypes

Col_0    datetime64[ns]
Col_1           float64
Col_2           float64
dtype: object

Map each color to every value present in the datetime column by defining a list of desired colors inside a dictionary.

c_dict = df['Col_0'].map(pd.Series(data=list('rgb'), index=df['Col_0'].values).to_dict())
df.plot.scatter(x='Col_1', y='Col_2', c=c_dict, alpha=0.8, title='Scatter-Plot')

It wouldn't be practical to use a list and populate colors for every value to be mapped against. In such cases, you are better of using colormaps to do the mapping.

c_dict = df['Col_0'].map(pd.Series(data=np.arange(3), index=df['Col_0'].values).to_dict())
df.plot.scatter(x='Col_1', y='Col_2', c=c_dict, title='Scatter-Plot', cmap=plt.cm.rainbow)

Collectives™ on Stack Overflow

Use of timestamp or datetime object for colorcoding of a scatterplot?

1 Answer 1

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

Comments

Your Answer

Sign up or log in

Post as a guest

Related