1

I am trying to plot a vertical line using axvline but ending up with the following error.

Update: the question deviates from posted references because only discreet x-axis tick locations are used (as stated by Trenton McKinney).

Error

TypeError: '<' not supported between instances of 'Timestamp' and 'numpy.float64'

Code

import pandas as pd

df = pd.DataFrame({'A': {0: '2020-01-01 06:00:00', 
                    1: '2020-01-01 18:00:00', 
                    2: '2020-01-02 06:00:00',
                    3: '2020-01-02 18:00:00',
                    4: '2020-01-03 06:00:00',
                    5: '2020-01-03 18:00:00',
                    6: '2020-01-04 06:00:00',
                    7: '2020-01-04 18:00:00'},
              'B': {0: 5, 1: 5, 2: 6, 3:6, 4:7, 5:7, 6:1, 7:1}})

df['A'] = pd.to_datetime(df['A'])
df= df.set_index('A')
ax = df.plot.bar()
ax.axvline(pd.to_datetime('2020-01-02 18:00:00'), color='grey', zorder=1, linestyle='--', marker="v" ) 
ax.axvline(pd.to_datetime('2020-01-03 00:00:00'), color='grey', zorder=1, linestyle='--', marker="v" )
ZF007
  • 3,708
  • 8
  • 29
  • 48
user1703276
  • 353
  • 1
  • 4
  • 14
  • In matplotlib the bars are located in integer xticks (0, 1, 2...), try `ax.axvline(3)` – RichieV Aug 12 '20 at 03:12
  • I need to use the timestamp to plot the lines. – user1703276 Aug 12 '20 at 03:37
  • 1
    Does this answer your question? [How do you plot a vertical line on a time series plot in Pandas?](https://stackoverflow.com/questions/19213789/how-do-you-plot-a-vertical-line-on-a-time-series-plot-in-pandas) – Daniel Widdis Aug 12 '20 at 20:43
  • 1
    The issue is, a bar is plotted on a bar plot based upon an integer location beginning at 0. The x-axis tick locations are discreet (i.e. 0, 1, 2, ..., n-1), so there are no tick locations in between. The dates are only text labels. – Trenton McKinney Aug 12 '20 at 20:48

1 Answers1

2

You could build your own barchart with a true datetime x-axis using ax.vlines:

import pandas as pd
from matplotlib.dates import DateFormatter

df = pd.DataFrame({'A': {0: '2020-01-01 06:00:00', 
                    1: '2020-01-01 18:00:00', 
                    2: '2020-01-02 06:00:00',
                    3: '2020-01-02 18:00:00',
                    4: '2020-01-03 06:00:00',
                    5: '2020-01-03 18:00:00',
                    6: '2020-01-04 06:00:00',
                    7: '2020-01-04 18:00:00'},
              'B': {0: 5, 1: 5, 2: 6, 3:6, 4:7, 5:7, 6:1, 7:1}})

df['A'] = pd.to_datetime(df['A'])
df= df.set_index('A')
fig, ax = plt.subplots()
ax.vlines(df.index, [0]*df.shape[0], df['B'], lw=25)
ax.axvline(pd.Timestamp('2020-01-02 18:00:00'), color='grey', zorder=1, linestyle='--', marker="v" ) 
ax.axvline(pd.Timestamp('2020-01-03 00:00:00'), color='grey', zorder=1, linestyle='--', marker="v" )
myFmt = DateFormatter("%m/%d/%Y %H:%M:%S")
ax.xaxis.set_major_formatter(myFmt)
ax.set_xticks(df.index)
for label in ax.xaxis.get_ticklabels():
    label.set_rotation(90)
ax.set_ylim(0,);

enter image description here

Scott Boston
  • 147,308
  • 15
  • 139
  • 187