如何使用 matplotlib 或 plotly 放大 Python 中的图形？

Question

I want to zoom in to the red section of the graph. I have the following code snippet. What changes do I need to make in my code either using matplotlib or plotly to achieve the desired results. Please prefer plotly library if you can.

# Visualize decision tree predictions

predictions = treePrediction
valid = df[x.shape[0]:]
valid["Predictions"] = predictions
plt.figure(figsize=(12, 7))
plt.title("Apple's Stock Price Prediction Model(Decision Tree Regressor Model)")
plt.xlabel("Days")
plt.ylabel("Close Price USD ($)")
plt.plot(df["Mean"])
plt.plot(valid[["Mean", "Predictions"]])
plt.legend(["Original", "Valid", "Predictions"])
plt.show()

Answer 1

Referring to the published node book, I have reproduced the graph in question. Since the nodebook was using locale CSV data, I retrieved the stock price data from yfinance and replaced the closing price with df['Mean'] . Plotly, a graph_object was used to add each graph. I have also added a button to select the period for zooming. See this page for details . We are adding 3 months as a period selection.

import plotly.graph_objects as go

fig = go.Figure()
fig.add_trace(go.Scatter(mode='lines', x=df.index, y=df['Mean'], line_color='blue', name='Original'))
fig.add_trace(go.Scatter(mode='lines', x=valid.index, y=valid['Mean'], line_color='orange', name='Valid'))
fig.add_trace(go.Scatter(mode='lines', x=valid.index, y=valid['Predictions'], line_color='green',name='Predictions'))

fig.update_layout(
    autosize=True,
    height=600,
    title="Apple's Stock Price Prediction Model(Decision Tree Regressor Model)",
    xaxis_title="Days",
    yaxis_title="Close Price USD ($)",
    template='plotly_white'
)

# Add range slider
fig.update_layout(
    xaxis=dict(
        rangeselector=dict(
            buttons=list([
                dict(count=1,
                     label="1m",
                     step="month",
                     stepmode="backward"),
                dict(count=3,
                     label="3m",
                     step="month",
                     stepmode="backward"),              
                dict(count=6,
                     label="6m",
                     step="month",
                     stepmode="backward"),
                dict(count=1,
                     label="YTD",
                     step="year",
                     stepmode="todate"),
                dict(count=1,
                     label="1y",
                     step="year",
                     stepmode="backward"),
                dict(step="all")
            ])
        ),
        rangeslider=dict(
            visible=True
        ),
        type="date"
    )
)

fig.show()

Select the most recent month using the Select Period button.

The range slider at the bottom of the graph can be used to select any range.