Plot line segments between two dates in matplotlib

Question

Given the following data:

dt = pd.DataFrame.from_dict(
    {
        "thing": {0: "A", 1: "B", 2: "C"},
        "min": {
            0: "2021-11-01 00:00:00+00:00",
            1: "2021-11-01 00:00:00+00:00",
            2: "2021-11-01 00:00:00+00:00",
        },
        "max": {
            0: "2021-11-02 00:00:00+00:00",
            1: "2021-11-05 00:00:00+00:00",
            2: "2021-11-07 00:00:00+00:00",
        },
    }
).assign(
    min=lambda x: pd.to_datetime(x["min"]),
    max=lambda x: pd.to_datetime(x["max"]),
)

Which looks like:

|    | thing   | min                       | max                       |
|---:|:--------|:--------------------------|:--------------------------|
|  0 | A       | 2021-11-01 00:00:00+00:00 | 2021-11-02 00:00:00+00:00 |
|  1 | B       | 2021-11-01 00:00:00+00:00 | 2021-11-05 00:00:00+00:00 |
|  2 | C       | 2021-11-01 00:00:00+00:00 | 2021-11-07 00:00:00+00:00 |

I would like to create a plot which has thing on the y-axis, and each a line representing the min / max on the x-axis.

Eg:

So the x-axis is the date, and the y-axis represents each 'thing'.

mozway · Accepted Answer · 2023-01-31 14:41:14Z

2

One option first reshaping with melt:

import matplotlib.pyplot as plt

ax = plt.subplot()
for k, g in dt.reset_index().melt(['index', 'thing']).groupby('thing'):
    g.plot(x='value', y='index', ax=ax, label=k, marker='s')

labels = dt['thing'].unique()

ax.set_yticks(range(len(labels)), labels)
ax.invert_yaxis()

Output:

alternative with seaborn

import seaborn as sns

g = sns.relplot(data=dt.reset_index().melt(['index', 'thing']),
                x='value', y='index', hue='thing',
                kind='line', marker='s')
labels = dt['thing'].unique()

g.ax.set_yticks(range(len(labels)), labels)
g.ax.invert_yaxis()

g.set_xticklabels(rotation=30)

Output:

pure matplotlib

ax = plt.subplot()
for row in dt.itertuples():
    ax.plot((row.min, row.max), (row.Index,)*2, label=row.thing, marker='s')

ax.set_yticks(range(len(dt)), dt['thing'])

ax.invert_yaxis()

If you want to keep the original order of the rows:

for row in dt.reset_index(drop=True).itertuples():
    ax.plot((row.min, row.max), (row.Index,)*2, label=row.thing, marker='s')

# OR

for i, row in enumerate(dt.itertuples()):
    ax.plot((row.min, row.max), (i,)*2, label=row.thing, marker='s')

edited Jan 31, 2023 at 14:41

answered Jan 31, 2023 at 14:18

mozway

267k13 gold badges56 silver badges106 bronze badges

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

6 Comments

baxx Over a year ago

thanks - for the sns version as well, is there an approach using default mpl for the plotting instead of df.plot (or sns) though ? I tried ax.plot with the melt approach and it didn't work

mozway Over a year ago

Sure, see last update

baxx Over a year ago

Note - this is also dependent on the labels being in an ordering with sorted - if the labels were DBC instead of ABC the approach would give the wrong result

mozway Over a year ago

The ordering relies on the index, you can use enumerate if you prefer, or change the index. You haven't given any specifics for that so I assumed ;)

mozway Over a year ago

Yes, the sorted should be removed

|

Collectives™ on Stack Overflow

Plot line segments between two dates in matplotlib

1 Answer 1

alternative with seaborn

pure matplotlib

6 Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

alternative with seaborn

pure matplotlib

6 Comments

Your Answer

Sign up or log in

Post as a guest

Related