匡醍量化|大富翁量化

Pandas Styler & Plotting: Data Visualization for Quants

中文 📅 2025-04-03 👁 views this month —

Pandas DataFrames offer robust styling capabilities, enabling Excel-like conditional formatting via the Styler object. Additionally, Pandas’ built-in plotting methods support a wide range of chart types, effortlessly meeting data visualization requirements.

1. Tables and Styling

Pandas DataFrames provide powerful styling features, allowing for Excel-like conditional coloring through the Styler object. Below are key methods and examples:

1.1. Basic Styling

Access styling features via DataFrame.style, which supports chained method calls:

df.style.set_caption("标题").set_properties(**{'background-color': 'lightgray'})

50%

1.2. Conditional Coloring

1.2.1. Single-Column Conditional Coloring

def color_negative_red(val):
    color = 'red' if val < 0.2 else 'black'
    return f'color: {color}'
df.style.applymap(color_negative_red)

50%

1.2.2. Multi-Column Conditional Coloring

df.style.apply(lambda x: ['background: yellow' if v > 0.2 else '' for v in x], 
                        subset=['A', 'C'])

50%

1.2.3. Highlighting Extremes

df.style.highlight_max(color='lightgreen').highlight_min(color='pink')

50%

1.2.4. Gradient Backgrounds

df.style.background_gradient(cmap='Blues', subset=['B'])

50%

1.2.5. Bar Chart Styling

df.style.bar(subset=['C'], color='#5fba7d')

50%

1.2.6. Custom Table Styles

headers = {'selector': 'th',
    'props': 'background-color: #5e17eb; color: white;'}
df.style.set_table_styles([headers])

50%

1.2.7. Dynamic Conditional Coloring (Complex Logic)

def highlight_risk(row):
    # 当A列>90且B列<50时标黄
    return ['background: yellow' if (row['A']>0.3) & 
    (row['B']<0.5) else '' for _ in row]  # 返回与行等长的样式列表
df.style.apply(highlight_risk, axis=1)  # axis=1表示按行处理

50%

Notes

- Styling only applies in Jupyter Notebooks or when exported to HTML; it does not modify the underlying data. - Use the `subset` parameter to restrict the coloring scope. - Gradient backgrounds (`background_gradient`) allow adjustment of the color scale range (`low=0.2, high=0.8`).
---

2. Pandas Built-in Plotting

In Pandas, we often deal with multi-column data along with row and column labels. Pandas includes built-in methods to simplify plotting from DataFrames and Series.

2.1. Line Charts

Both Series and DataFrame objects have a plot attribute for creating basic charts. By default, plot() generates line charts.

s = pd.Series(np.random.standard_normal(10).cumsum(), index=np.arange(0, 100, 10))
s.plot()

50%

The index of the Series object is passed to Matplotlib and used to plot the x-axis. You can disable index usage with use_index=False. X-axis ticks and limits can be adjusted via xticks and xlim, while the y-axis is controlled by yticks and ylim. A partial list of plot parameters is shown in the table below:

Parameter Description
alpha Transparency of the plot fill (between 0 and 1)
ax Matplotlib Axes object; defaults to the current Axes (gca())
colormap Specifies the color map (e.g., 'viridis')
figsize Image size in format (width, height) (in inches)
fontsize Font size for tick labels
grid Whether to display grid lines (defaults to None, following Matplotlib’s default style)
kind Plot type: 'line' (line chart, default), 'bar' (bar chart), 'barh' (horizontal bar chart), 'hist' (histogram), 'box' (box plot), 'kde'/'density' (kernel density estimate), 'area' (area chart), 'pie' (pie chart)
label Legend label name
legend Whether to display the legend (default is False)
logx/logy Whether to use logarithmic scales for the x/y axis (default is False)
loglog Whether to use logarithmic scales for both x and y axes
position Position of bars in bar charts (avoid conflict with kind='bar')
rot Rotation angle for tick labels (e.g., 45 for 45 degrees)
secondary_y Whether to use a second y-axis on the right (default is False)
style Line style (e.g., 'k--' for black dashed line)
table Whether to display a data table below the chart (default is False)
title Chart title (string)
use_index Whether to use the Series index as x-axis tick labels (default is True)
xerr/yerr Add error bars to bar charts
xlim/ylim Set x/y-axis display range (format: (min, max))
xticks/yticks Customize x/y-axis tick values (list)
**kwds Other Matplotlib plotting parameters (e.g., color='red')

Most Pandas plotting methods accept an optional ax parameter, which can be a Matplotlib subplot object, allowing for more flexible positioning of subplots within a grid layout.

The DataFrame’s plot method draws each column as a line in the same subplot and automatically creates a legend.

df = pd.DataFrame(np.random.standard_normal((10, 4)).cumsum(0),
                  columns=['A', 'B', 'C', 'D'],
                  index=np.arange(0, 100, 10))
plt.style.use('grayscale')
df.plot()

50%

Notes

Here, `plt.style.use('grayscale')` is used to set the color scheme to grayscale.
The `plot` attribute includes many methods for different plot types. For example, `df.plot()` is equivalent to `df.plot.line()`.

Notes

Additional keyword arguments passed to `plot` are forwarded to the corresponding Matplotlib plotting function. Therefore, further customization of charts requires deeper knowledge of the Matplotlib API.
DataFrames also offer flexible options for handling columns, such as plotting all columns in the same subplot or creating separate subplots for each. The table below shows DataFrame-specific `plot` parameters:
Parameter Description
subplots Whether to create subplots for each column (default is False)
sharex If subplots=True, whether to share the x-axis (default is True when ax=None)
sharey If subplots=True, whether to share the y-axis (default is False)
layout Subplot row/column layout in format (rows, columns)
legend Add legends to subplots (default is True)
sort_columns Whether to sort columns by name (default is False)

2.2. Bar Charts

plot.bar() and plot.barh() are used to draw vertical and horizontal bar charts, respectively. For bar charts, the index of the Series or DataFrame is used as the tick marks on the x-axis (for bar) or y-axis (for barh).

fig, axes = plt.subplots(2,1)
data = pd.Series(np.random.uniform(size=16), index=list('abcdefghijklmnop'))
data.plot.bar(ax=axes[0], color='k', alpha=0.7)
data.plot.barh(ax=axes[1], color='k', alpha=0.7)

50%

For DataFrames, bar charts group the values of each row and display them side by side.

df = pd.DataFrame(np.random.uniform(size=(6, 4)),
        index=["one", "two", "three", "four", "five", "six"],
        columns=pd.Index(["A", "B", "C", "D"], name="Genus"))
df.plot.bar()

50%

Note that the column name "Genus" is used as the legend title.

Passing stacked=True generates a stacked bar chart for the DataFrame, where the values of each row are stacked horizontally.

df.plot.bar(stacked=True,alpha=0.5)

50%