How to put the xticks equal to this figure

Question

How to put the xticks equal to this figure

Navigation

#1 by (0 votes)

1

I'm doing a Python course for finance. Hence in an exercise you are asked to plot a chart using the DataFrames. So far so good, I have managed to do everything right, but my xticks do not match those of the result!

I basically have to get this:

But I'm getting it here:

Would anyone know how to get the same xticks as the citizen?

Note: It is time series in Pandas that should be plotted ford, tesla and gm are DataFrames and I am plotting the 'Open' column of this dataframe. The DataFrames indexes are all the same, dates from 2012-01-01 to 2017-01-01.

obs2 .: The code used by prof. was as follows:

tesla['Open'].plot(label='Tesla',figsize=(16,8),title='Open Price')
gm['Open'].plot(label='GM')
ford['Open'].plot(label='Ford')
plt.legend()

python-3.x matplotlib pandas

asked by anonymous 28.06.2018 / 22:33

1 answer

Array inside Array use .htacces properly for links

score 0 · Accepted Answer

Hello

First, there are several different ways to resolve these customization issues. Below is a form I usually use and it seems more readable to me.

In this case, I set a axis where all the plots and customizations will be set:

# Definição do axis
fig, ax = plt.subplots(1,1)

# Note que agora é necessário definir o axis que cada plot será exibido
tesla['Open'].plot(ax = ax, label = "Tesla", figsize = (16,8), title = "Open Price")
gm['Open'].plot(ax = ax, label = "GM")
ford['Open'].plot(ax = ax, label = "Ford")

By the image of your question, it seems to me that xticks has been separated in 6-month intervals. Considering that the indexes of all DataFrames are the same, it is necessary to define the position where each xtick will appear and then format so that 2016-07 appears, for example.

# Cria um vetor com as datas a cada 6 meses
# O parâmetro freq pode receber diversas frequências diferentes (Y, m, d, ...)
positions = pd.date_range(tesla.index[0], tesla.index[-1], freq='6m')

# Formata os labels
labels = positions.strftime('%Y-%m')

Now let's set the formatted dates on the abscissa axis

# Remove localizadores desnecessários no eixo x
ax.xaxis.set_minor_locator(plt.NullLocator())

# Seta a posição das datas
ax.set_xticks(positions[1:])
# Seta os labels nas devidas posições
ax.set_xticklabels(labels[1:])

To tilt the labels, simply add a rotation to the xticks with the command plt.xticks(rotation = angle) , where angle is the angle of rotation in degrees of the xticks with respect to the axis of the abscissa.

plt.xticks(rotation = 30)

Finally, we set the graphic to fit the image

plt.autoscale(axis = 'x', tight=True)

Complete code:

fig, ax = plt.subplots(1,1)

tesla['Open'].plot(ax = ax, label = "Tesla", figsize = (16,8), title = "Open Price")
gm['Open'].plot(ax = ax, label = "GM")
ford['Open'].plot(ax = ax, label = "Ford")

positions = pd.date_range(tesla.index[0], tesla.index[-1], freq='6m')
labels = positions.strftime('%Y-%m')

ax.xaxis.set_minor_locator(plt.NullLocator())
ax.set_xticks(positions[1:])
ax.set_xticklabels(labels[1:])

plt.xticks(rotation = 30)

plt.legend()
plt.autoscale(axis = 'x', tight=True)

References:

a>

link