Pyplot ticks disappear when formatting time

问题

I am using matplotlib to plot a serie of values per week.
The dataframe I create (see code below) looks like this:

Dataframe: df_dates   

| Name_Of_index |       |  
|------------   |----   |  
| 2017-01-22    | 13    |  
| 2017-01-29    | 0     |  
| 2017-02-05    | 5     |  
| 2017-02-12    | 37    |    

Freq: W-SUN, dtype: int64

My code is supposed to draw a bar plot for number per week/timestamp.

def plot_contact_with_waters_per_week(dataframe, target_name,contact_with_water_value='contact_with_water', date_col='TBL_DAT'):
   df_contact_with_waters=dataframe[dataframe[target_name]==contact_with_water_value]
   contact_with_waters_per_week=pd.Series((df_contact_with_waters[target_name].values),index=df_contact_with_waters[date_col]).resample(rule='W').count()
   axis_now=contact_with_waters_per_week.plot(kind='bar')
   axis_now.set_ylabel("Total number of contact_with_waters per week")
   axis_now.xaxis_date() 
   axis_now.xaxis.set_major_locator(mdates.WeekdayLocator(byweekday=SU))
   axis_now.xaxis.set_major_formatter(mdates.DateFormatter('%b %d'))
   plt.show()

The thing that bothers me is that as soon as I format the x-axis, all ticks become invisible.

Now, if I just delete the three xaxis lines:

   axis_now.xaxis_date() 
   axis_now.xaxis.set_major_locator(mdates.WeekdayLocator(byweekday=SU))
   axis_now.xaxis.set_major_formatter(mdates.DateFormatter('%b %d'))

I do get my plot with timestamps, but they have the full length and do not look pretty:

Now, I suppose I make a mistake in formatting my timestamps, can you please have a look and give me some advice? It would be much appreciated.

回答1:

In case anybody ever runs into the same issue, here is the code that worked:

def plot_contact_with_waters_per_week(dataframe, target_name,contact_with_water_value='contact_with_water', date_col='TBL_DAT'):
   df_contact_with_waters=dataframe[dataframe[target_name]==contact_with_water_value]
   contact_with_waters_per_week=pd.Series((df_contact_with_waters[target_name].values),index=df_contact_with_waters[date_col]).resample(rule='W').count()
   fig, ax = plt.subplots()
   plt.bar(height=contact_with_waters_per_week,left=contact_with_waters_per_week.index,width=3)
   plt.ylabel("Total number of contact_with_waters per week")
   ax.xaxis_date() 
   ax.xaxis.set_major_locator(mdates.WeekdayLocator(byweekday=SU))
   ax.xaxis.set_major_formatter(mdates.DateFormatter('%b %d'))
   my_xticks = ax.get_xticks()
   my_xticks_biweekly=my_xticks[::2].copy()
   plt.xticks(my_xticks_biweekly,visible=True, rotation="vertical")
   plt.show()

A big thank you to ImportanceOfBeingErnest!

来源：https://stackoverflow.com/questions/48804513/pyplot-ticks-disappear-when-formatting-time

标签

python

matplotlib

graph

visualization