I have a pandas dataframe with thee columns 'SystemState_InitDate', 'SystemState_FinalDate' and 'State_Desc'.

I want to plot a chronogram. To do that, I have generated the following function

for idx, row in df_cleaned.iterrows():
    ax.barh(y=row['State_Desc'], left=row['SystemState_InitDate'], width=row['SystemState_FinalDate'] - row['SystemState_InitDate'], color=colors[row['State_Desc']], edgecolor='black')

ax.set_xlim([df_cleaned['SystemState_InitDate'].min(), df_cleaned['SystemState_FinalDate'].max()])

ax.xaxis.set_major_locator(mdates.HourLocator(interval=6))
ax.xaxis.set_major_formatter(mdates.DateFormatter('%Y-%m-%d %H:%M'))

and later on I plot as

    plt.xlabel('Fecha y Hora')
plt.ylabel('Estado del Sistema')
plt.title('Cronograma de Estados del Sistema')

The resulting plot, as you can see in the figure, has a wide blank space in the middle, since nothing was done during those times.

I was wondering if there was a way for me to cut all the times in which nothing is happening so that I can get a more dense figure where the details can be appreciated better without that useless white space[![enter image description here][1]][1].

I have tried chatgpt but it has suggested me to use the following:

  prev_final_date = None
    for idx, row in df_cleaned.iterrows():
        if prev_final_date is not None and row['SystemState_InitDate'] > prev_final_date:
            # Si hay un hueco, no dibujar nada en ese hueco
            continue
        ax.barh(y=row['State_Desc'], left=row['SystemState_InitDate'], width=row['SystemState_FinalDate'] - row['SystemState_InitDate'], color=colors.get(row['State_Desc'], 'grey'), edgecolor='black')
        prev_final_date = row['SystemState_FinalDate']
    
    ax.set_xlim([tiempoInicial, tiempoFinal])

    ax.xaxis.set_major_locator(mdates.HourLocator(interval=6))
    ax.xaxis.set_major_formatter(mdates.DateFormatter('%Y-%m-%d %H:%M'))

which returns a blank image.

Can someone please tell me how can I solve my problem? Any help is appreciated.
[1]: https://i.sstatic.net/pBbPhysf.png