Plot dataframe with datetime index

Question

            is_avail   valu data_source
2015-08-07     False  0.282    source_a
2015-08-23     False  0.296    source_a
2015-09-08     False  0.433    source_a
2015-10-01      True  0.169    source_a
2015-10-10      True  0.160    source_a
2015-11-02     False  0.179    source_a
2016-03-09     False  0.557    source_a
2016-04-26     False  0.770    source_a
2016-05-05     False  0.826    source_a
2016-05-12     False  0.826    source_a
2016-05-28     False  0.747    source_a
2016-06-06     False  0.796    source_a
2016-07-31     False  0.322    source_a
2016-08-25      True  0.136    source_a
2016-09-10     False  0.180    source_a
2016-11-13     False  0.492    source_a
2016-12-15      True  0.124    source_a
2016-12-31     False  0.533    source_a
2017-03-28     False  0.524    source_a
2015-06-27      True  0.038    source_b
2015-07-30      True  0.035    source_b
2015-08-06     False  0.205    source_b
2015-08-09     False  0.241    source_b
2015-08-16      True  0.025    source_b
2015-08-19      True  0.092    source_b
2015-08-26     False  0.264    source_b
2015-08-29     False  0.312    source_b

I want to plot the valu column in the dataframe above, with separate line plots for source_a rows and source_b rows and separate colors when is_avail is True or False. Currently, I am doing this:

df['valu'].plot()

However, this shows everything as a single line plot. How to get to the objective above?

mechanical_meat · Accepted Answer · 2017-05-05 17:20:35Z

5

It's not so simple of a solution but this works afaict.

import pandas as pd
import matplotlib.pyplot as plt
import matplotlib.dates as mdates 

df = df.sort_index()

fig, ax = plt.subplots()
for k,g in df.groupby(['is_avail','data_source'])['valu']:
    ax.plot_date(pd.to_datetime(g.index),g,'v-',label="{}, {}".format(*k))

ax.xaxis.set_major_locator(mdates.MonthLocator(interval=3))
ax.xaxis.set_major_formatter(mdates.DateFormatter('%b\n%Y'))
ax.legend()
plt.show()

Result:

edited May 5, 2017 at 17:20

answered May 5, 2017 at 16:12

mechanical_meat

170k25 gold badges237 silver badges231 bronze badges

Sign up to request clarification or add additional context in comments.

Comments

VinceP · Accepted Answer · 2017-05-05 16:57:00Z

3

How about this? Assuming your data is in the dataframe df:

import pandas as pd
df.is_avail = df.is_avail.astype('str')
grouped = df.groupby(['data_source','is_avail'])

fig, ax = plt.subplots()
for lab, gr in grouped:
    gr.valu.plot(ax=ax, label = lab)
plt.legend()

answered May 5, 2017 at 16:57

VinceP

2,1832 gold badges23 silver badges31 bronze badges

Collectives™ on Stack Overflow

Plot dataframe with datetime index

2 Answers 2

Comments

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

Comments

Comments

Your Answer

Sign up or log in

Post as a guest

Related