Plotting datetime for several years but showing only twelve months on x-axis

Question

I am trying to plot data with datetime as x-axis. The data is collected over several years. I can convert the Date to datetime format, extract the year portion of the datetime and plot using that as color.

df['Date']=pd.to_datetime(df['Date'],errors='coerce') 
df['year']=df['Date'].dt.year 
df['mthday']=df['Date'].dt.strftime('%d-%b') 
df=df.sort_values(by='Date') 
plot=px.line(df,x='Date',y='value',color='year')

(I have tried the solution at this link: Plotly: How to plot just month and day on x axis? (Ignore year))

This gives me the graph below. The different years are correctly shown in different colors. However, the x-axis spans from 2014 to 2020. I want the x-axis to only show Jan to Dec and the graphs for different years to be superposed on this. (aka: shortening/ collapsing the x-axis to only one set of months from Jan to Dec)

Edited to add sample from df:

    Date/   value/value2/value3/Note/Unnamed/year/mthday
187 2015-11-10  125 73  79.0    NaN NaN 2015    10-Nov
108 2015-01-25  132 88  61.0    NaN NaN 2015    25-Jan
85  2014-12-25  138 86  69.0    NaN NaN 2014    25-Dec
154 2015-07-24  131 80  62.0    NaN NaN 2015    24-Jul
226 2016-09-18  120 73  61.0    NaN NaN 2016    18-Sep
261 2018-06-07  135 72  64.0    NaN NaN 2018    07-Jun
59  2014-11-20  158 79  70.0    NaN NaN 2014    20-Nov
118 2015-02-12  145 77  70.0    NaN NaN 2015    12-Feb
287 2019-09-02  130 79  66.0    NaN NaN 2019    02-Sep
228 2016-09-26  104 76  59.0    NaN NaN 2016    26-Sep

(Actually for the graph, I would want to plot value,value2,value3 for all the years. In this example, I have only tried for value for all the years)

2nd edit for possible answer: Hi, I think i have found a way to do this. I use the mthday column and convert it back to datetime. It defaults the year value to 1900. I then plot this using mthday as x-axis and year as color. Then hide the year value in the x-axis tickmark.

df['mthday']=pd.to_datetime(df['mthday'],format='%d-%b',errors='coerce')
fig=px.line(df,x='mthday',y='value',color='year')
fig.update_layout(xaxis_tickformat='%d-%b')

Let me know if you have a better way to accomplish this.

Answer 1

Posting below as one way that I have found to do this: I use the mthday column and convert it back to datetime type. This puts all the year values for all the rows to default value of 1900. I then use mthday as x-axis and hide the year by specifying the x-axis tickmarks.

df['mthday']=df['Date'].dt.strftime('%d-%b')
df['mthday']=pd.to_datetime(df['mthday'],format='%d-%b',errors='coerce')
df=df.sort_values(by='Date') 
plot=px.line(df,x='mthday',y='value',color='year')
plot.update_layout(xaxis_tickformat='%d-%b')