Visualization

We use the standard convention for referencing the matplotlib API:

In [1]: import matplotlib.pyplot as plt

The plots in this document are made using matplotlib’s ggplot style (new in version 1.4):

import matplotlib
matplotlib.style.use('ggplot')

We provide the basics in pandas to easily create decent looking plots. See the ecosystem section for visualization libraries that go beyond the basics documented here.

Note

All calls to np.random are seeded with 123456.

Basic Plotting: plot

See the cookbook for some advanced strategies

The plot method on Series and DataFrame is just a simple wrapper around plt.plot():

In [2]: ts = pd.Series(np.random.randn(1000), index=pd.date_range('1/1/2000', periods=1000))

In [3]: ts = ts.cumsum()

In [4]: ts.plot()
Out[4]: <matplotlib.axes._subplots.AxesSubplot at 0x132cd4850>
_images/series_plot_basic.png

If the index consists of dates, it calls gcf().autofmt_xdate() to try to format the x-axis nicely as per above.

On DataFrame, plot() is a convenience to plot all of the columns with labels:

In [5]: df = pd.DataFrame(np.random.randn(1000, 4), index=ts.index, columns=list('ABCD'))

In [6]: df = df.cumsum()

In [7]: plt.figure(); df.plot();
_images/frame_plot_basic.png

You can plot one column versus another using the x and y keywords in plot():

In [8]: df3 = pd.DataFrame(np.random.randn(1000, 2), columns=['B', 'C']).cumsum()

In [9]: df3['A'] = pd.Series(list(range(len(df))))

In [10]: df3.plot(x='A', y='B')
Out[10]: <matplotlib.axes._subplots.AxesSubplot at 0x135189110>
_images/df_plot_xy.png

Note

For more formatting and styling options, see below.

Other Plots

Plotting methods allow for a handful of plot styles other than the default Line plot. These methods can be provided as the kind keyword argument to plot(). These include:

For example, a bar plot can be created the following way:

In [11]: plt.figure();

In [12]: df.ix[5].plot(kind='bar'); plt.axhline(0, color='k')
Out[12]: <matplotlib.lines.Line2D at 0x1346eb610>
_images/bar_plot_ex.png

New in version 0.17.0.

You can also create these other plots using the methods DataFrame.plot.<kind> instead of providing the kind keyword argument. This makes it easier to discover plot methods and the specific arguments they use:

In [13]: df = pd.DataFrame()

In [14]: df.plot.<TAB>
df.plot.area     df.plot.barh     df.plot.density  df.plot.hist     df.plot.line     df.plot.scatter
df.plot.bar      df.plot.box      df.plot.hexbin   df.plot.kde      df.plot.pie

In addition to these kind s, there are the DataFrame.hist(), and DataFrame.boxplot() methods, which use a separate interface.

Finally, there are several plotting functions in pandas.tools.plotting that take a Series or DataFrame as an argument. These include

Plots may also be adorned with errorbars or tables.

Bar plots

For labeled, non-time series data, you may wish to produce a bar plot:

In [15]: plt.figure();

In [16]: df.ix[5].plot.bar(); plt.axhline(0, color='k')
Out[16]: <matplotlib.lines.Line2D at 0x133e28410>
_images/bar_plot_ex.png

Calling a DataFrame’s plot.bar() method produces a multiple bar plot:

In [17]: df2 = pd.DataFrame(np.random.rand(10, 4), columns=['a', 'b', 'c', 'd'])

In [18]: df2.plot.bar();
_images/bar_plot_multi_ex.png

To produce a stacked bar plot, pass stacked=True:

In [19]: df2.plot.bar(stacked=True);
_images/bar_plot_stacked_ex.png

To get horizontal bar plots, use the barh method:

In [20]: df2.plot.barh(stacked=True);
_images/barh_plot_stacked_ex.png

Histograms

New in version 0.15.0.

Histogram can be drawn by using the DataFrame.plot.hist() and Series.plot.hist() methods.

In [21]: df4 = pd.DataFrame({'a': np.random.randn(1000) + 1, 'b': np.random.randn(1000),
   ....:                     'c': np.random.randn(1000) - 1}, columns=['a', 'b', 'c'])
   ....: 

In [22]: plt.figure();

In [23]: df4.plot.hist(alpha=0.5)
Out[23]: <matplotlib.axes._subplots.AxesSubplot at 0x132832150>
_images/hist_new.png

Histogram can be stacked by stacked=True. Bin size can be changed by bins keyword.

In [24]: plt.figure();

In [25]: df4.plot.hist(stacked=True, bins=20)
Out[25]: <matplotlib.axes._subplots.AxesSubplot at 0x132817850>
_images/hist_new_stacked.png

You can pass other keywords supported by matplotlib hist. For example, horizontal and cumulative histgram can be drawn by orientation='horizontal' and cumulative='True'.

In [26]: plt.figure();

In [27]: df4['a'].plot.hist(orientation='horizontal', cumulative=True)
Out[27]: <matplotlib.axes._subplots.AxesSubplot at 0x132b8ca90>
_images/hist_new_kwargs.png

See the hist method and the matplotlib hist documentation for more.

The existing interface DataFrame.hist to plot histogram still can be used.

In [28]: plt.figure();

In [29]: df['A'].diff().hist()
Out[29]: <matplotlib.axes._subplots.AxesSubplot at 0x132d97450>
_images/hist_plot_ex.png

DataFrame.hist() plots the histograms of the columns on multiple subplots:

In [30]: plt.figure()
Out[30]: <matplotlib.figure.Figure at 0x132def650>

In [31]: df.diff().hist(color='k', alpha=0.5, bins=50)
Out[31]: 
array([[<matplotlib.axes._subplots.AxesSubplot object at 0x135936bd0>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x133d28410>],
       [<matplotlib.axes._subplots.AxesSubplot object at 0x139948490>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x13052c610>]], dtype=object)
_images/frame_hist_ex.png

New in version 0.10.0.

The by keyword can be specified to plot grouped histograms:

In [32]: data = pd.Series(np.random.randn(1000))

In [33]: data.hist(by=np.random.randint(0, 4, 1000), figsize=(6, 4))
Out[33]: 
array([[<matplotlib.axes._subplots.AxesSubplot object at 0x129225a90>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x132d20d50>],
       [<matplotlib.axes._subplots.AxesSubplot object at 0x129842c50>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x122103350>]], dtype=object)
_images/grouped_hist.png

Box Plots

New in version 0.15.0.

Boxplot can be drawn calling Series.plot.box() and DataFrame.plot.box(), or DataFrame.boxplot() to visualize the distribution of values within each column.

For instance, here is a boxplot representing five trials of 10 observations of a uniform random variable on [0,1).

In [34]: df = pd.DataFrame(np.random.rand(10, 5), columns=['A', 'B', 'C', 'D', 'E'])

In [35]: df.plot.box()
Out[35]: <matplotlib.axes._subplots.AxesSubplot at 0x11d289150>
_images/box_plot_new.png

Boxplot can be colorized by passing color keyword. You can pass a dict whose keys are boxes, whiskers, medians and caps. If some keys are missing in the dict, default colors are used for the corresponding artists. Also, boxplot has sym keyword to specify fliers style.

When you pass other type of arguments via color keyword, it will be directly passed to matplotlib for all the boxes, whiskers, medians and caps colorization.

The colors are applied to every boxes to be drawn. If you want more complicated colorization, you can get each drawn artists by passing return_type.

In [36]: color = dict(boxes='DarkGreen', whiskers='DarkOrange',
   ....:              medians='DarkBlue', caps='Gray')
   ....: 

In [37]: df.plot.box(color=color, sym='r+')
Out[37]: <matplotlib.axes._subplots.AxesSubplot at 0x130051d50>
_images/box_new_colorize.png

Also, you can pass other keywords supported by matplotlib boxplot. For example, horizontal and custom-positioned boxplot can be drawn by vert=False and positions keywords.

In [38]: df.plot.box(vert=False, positions=[1, 4, 5, 6, 8])
Out[38]: <matplotlib.axes._subplots.AxesSubplot at 0x129781750>
_images/box_new_kwargs.png

See the boxplot method and the matplotlib boxplot documentation for more.

The existing interface DataFrame.boxplot to plot boxplot still can be used.

In [39]: df = pd.DataFrame(np.random.rand(10,5))

In [40]: plt.figure();

In [41]: bp = df.boxplot()
_images/box_plot_ex.png

You can create a stratified boxplot using the by keyword argument to create groupings. For instance,

In [42]: df = pd.DataFrame(np.random.rand(10,2), columns=['Col1', 'Col2'] )

In [43]: df['X'] = pd.Series(['A','A','A','A','A','B','B','B','B','B'])

In [44]: plt.figure();

In [45]: bp = df.boxplot(by='X')
_images/box_plot_ex2.png

You can also pass a subset of columns to plot, as well as group by multiple columns:

In [46]: df = pd.DataFrame(np.random.rand(10,3), columns=['Col1', 'Col2', 'Col3'])

In [47]: df['X'] = pd.Series(['A','A','A','A','A','B','B','B','B','B'])

In [48]: df['Y'] = pd.Series(['A','B','A','B','A','B','A','B','A','B'])

In [49]: plt.figure();

In [50]: bp = df.boxplot(column=['Col1','Col2'], by=['X','Y'])
_images/box_plot_ex3.png

Basically, plot functions return matplotlib Axes as a return value. In boxplot, the return type can be changed by argument return_type, and whether the subplots is enabled (subplots=True in plot or by is specified in boxplot).

When subplots=False / by is None:

  • if return_type is 'dict', a dictionary containing the matplotlib Lines is returned. The keys are “boxes”, “caps”, “fliers”, “medians”, and “whiskers”.

    This is the default of boxplot in historical reason. Note that plot.box() returns Axes by default same as other plots.

  • if return_type is 'axes', a matplotlib Axes containing the boxplot is returned.

  • if return_type is 'both' a namedtuple containing the matplotlib Axes

    and matplotlib Lines is returned

When subplots=True / by is some column of the DataFrame:

  • A dict of return_type is returned, where the keys are the columns of the DataFrame. The plot has a facet for each column of the DataFrame, with a separate box for each value of by.

Finally, when calling boxplot on a Groupby object, a dict of return_type is returned, where the keys are the same as the Groupby object. The plot has a facet for each key, with each facet containing a box for each column of the DataFrame.

In [51]: np.random.seed(1234)

In [52]: df_box = pd.DataFrame(np.random.randn(50, 2))

In [53]: df_box['g'] = np.random.choice(['A', 'B'], size=50)

In [54]: df_box.loc[df_box['g'] == 'B', 1] += 3

In [55]: bp = df_box.boxplot(by='g')
_images/boxplot_groupby.png

Compare to:

In [56]: bp = df_box.groupby('g').boxplot()
_images/groupby_boxplot_vis.png

Area Plot

New in version 0.14.

You can create area plots with Series.plot.area() and DataFrame.plot.area(). Area plots are stacked by default. To produce stacked area plot, each column must be either all positive or all negative values.

When input data contains NaN, it will be automatically filled by 0. If you want to drop or fill by different values, use dataframe.dropna() or dataframe.fillna() before calling plot.

In [57]: df = pd.DataFrame(np.random.rand(10, 4), columns=['a', 'b', 'c', 'd'])

In [58]: df.plot.area();
_images/area_plot_stacked.png

To produce an unstacked plot, pass stacked=False. Alpha value is set to 0.5 unless otherwise specified:

In [59]: df.plot.area(stacked=False);
_images/area_plot_unstacked.png

Scatter Plot

New in version 0.13.

Scatter plot can be drawn by using the DataFrame.plot.scatter() method. Scatter plot requires numeric columns for x and y axis. These can be specified by x and y keywords each.

In [60]: df = pd.DataFrame(np.random.rand(50, 4), columns=['a', 'b', 'c', 'd'])

In [61]: df.plot.scatter(x='a', y='b');
_images/scatter_plot.png

To plot multiple column groups in a single axes, repeat plot method specifying target ax. It is recommended to specify color and label keywords to distinguish each groups.

In [62]: ax = df.plot.scatter(x='a', y='b', color='DarkBlue', label='Group 1');

In [63]: df.plot.scatter(x='c', y='d', color='DarkGreen', label='Group 2', ax=ax);
_images/scatter_plot_repeated.png

The keyword c may be given as the name of a column to provide colors for each point:

In [64]: df.plot.scatter(x='a', y='b', c='c', s=50);
_images/scatter_plot_colored.png

You can pass other keywords supported by matplotlib scatter. Below example shows a bubble chart using a dataframe column values as bubble size.

In [65]: df.plot.scatter(x='a', y='b', s=df['c']*200);
_images/scatter_plot_bubble.png

See the scatter method and the matplotlib scatter documentation for more.

Hexagonal Bin Plot

New in version 0.14.

You can create hexagonal bin plots with DataFrame.plot.hexbin(). Hexbin plots can be a useful alternative to scatter plots if your data are too dense to plot each point individually.

In [66]: df = pd.DataFrame(np.random.randn(1000, 2), columns=['a', 'b'])

In [67]: df['b'] = df['b'] + np.arange(1000)

In [68]: df.plot.hexbin(x='a', y='b', gridsize=25)
Out[68]: <matplotlib.axes._subplots.AxesSubplot at 0x1338c8750>
_images/hexbin_plot.png

A useful keyword argument is gridsize; it controls the number of hexagons in the x-direction, and defaults to 100. A larger gridsize means more, smaller bins.

By default, a histogram of the counts around each (x, y) point is computed. You can specify alternative aggregations by passing values to the C and reduce_C_function arguments. C specifies the value at each (x, y) point and reduce_C_function is a function of one argument that reduces all the values in a bin to a single number (e.g. mean, max, sum, std). In this example the positions are given by columns a and b, while the value is given by column z. The bins are aggregated with numpy’s max function.

In [69]: df = pd.DataFrame(np.random.randn(1000, 2), columns=['a', 'b'])

In [70]: df['b'] = df['b'] = df['b'] + np.arange(1000)

In [71]: df['z'] = np.random.uniform(0, 3, 1000)

In [72]: df.plot.hexbin(x='a', y='b', C='z', reduce_C_function=np.max,
   ....:         gridsize=25)
   ....: 
Out[72]: <matplotlib.axes._subplots.AxesSubplot at 0x127894650>
_images/hexbin_plot_agg.png

See the hexbin method and the matplotlib hexbin documentation for more.

Pie plot

New in version 0.14.

You can create a pie plot with DataFrame.plot.pie() or Series.plot.pie(). If your data includes any NaN, they will be automatically filled with 0. A ValueError will be raised if there are any negative values in your data.

In [73]: series = pd.Series(3 * np.random.rand(4), index=['a', 'b', 'c', 'd'], name='series')

In [74]: series.plot.pie(figsize=(6, 6))
Out[74]: <matplotlib.axes._subplots.AxesSubplot at 0x12c246810>
_images/series_pie_plot.png

For pie plots it’s best to use square figures, one’s with an equal aspect ratio. You can create the figure with equal width and height, or force the aspect ratio to be equal after plotting by calling ax.set_aspect('equal') on the returned axes object.

Note that pie plot with DataFrame requires that you either specify a target column by the y argument or subplots=True. When y is specified, pie plot of selected column will be drawn. If subplots=True is specified, pie plots for each column are drawn as subplots. A legend will be drawn in each pie plots by default; specify legend=False to hide it.

In [75]: df = pd.DataFrame(3 * np.random.rand(4, 2), index=['a', 'b', 'c', 'd'], columns=['x', 'y'])

In [76]: df.plot.pie(subplots=True, figsize=(8, 4))
Out[76]: 
array([<matplotlib.axes._subplots.AxesSubplot object at 0x133e48dd0>,
       <matplotlib.axes._subplots.AxesSubplot object at 0x130b72d90>], dtype=object)
_images/df_pie_plot.png

You can use the labels and colors keywords to specify the labels and colors of each wedge.

Warning

Most pandas plots use the the label and color arguments (note the lack of “s” on those). To be consistent with matplotlib.pyplot.pie() you must use labels and colors.

If you want to hide wedge labels, specify labels=None. If fontsize is specified, the value will be applied to wedge labels. Also, other keywords supported by matplotlib.pyplot.pie() can be used.

In [77]: series.plot.pie(labels=['AA', 'BB', 'CC', 'DD'], colors=['r', 'g', 'b', 'c'],
   ....:                 autopct='%.2f', fontsize=20, figsize=(6, 6))
   ....: 
Out[77]: <matplotlib.axes._subplots.AxesSubplot at 0x1238e4950>
_images/series_pie_plot_options.png

If you pass values whose sum total is less than 1.0, matplotlib draws a semicircle.

In [78]: series = pd.Series([0.1] * 4, index=['a', 'b', 'c', 'd'], name='series2')

In [79]: series.plot.pie(figsize=(6, 6))
Out[79]: <matplotlib.axes._subplots.AxesSubplot at 0x133e48b90>
_images/series_pie_plot_semi.png

See the matplotlib pie documentation for more.

Plotting with Missing Data

Pandas tries to be pragmatic about plotting DataFrames or Series that contain missing data. Missing values are dropped, left out, or filled depending on the plot type.

Plot Type NaN Handling
Line Leave gaps at NaNs
Line (stacked) Fill 0’s
Bar Fill 0’s
Scatter Drop NaNs
Histogram Drop NaNs (column-wise)
Box Drop NaNs (column-wise)
Area Fill 0’s
KDE Drop NaNs (column-wise)
Hexbin Drop NaNs
Pie Fill 0’s

If any of these defaults are not what you want, or if you want to be explicit about how missing values are handled, consider using fillna() or dropna() before plotting.

Plotting Tools

These functions can be imported from pandas.tools.plotting and take a Series or DataFrame as an argument.

Scatter Matrix Plot

New in version 0.7.3.

You can create a scatter plot matrix using the scatter_matrix method in pandas.tools.plotting:

In [80]: from pandas.tools.plotting import scatter_matrix

In [81]: df = pd.DataFrame(np.random.randn(1000, 4), columns=['a', 'b', 'c', 'd'])

In [82]: scatter_matrix(df, alpha=0.2, figsize=(6, 6), diagonal='kde')
Out[82]: 
array([[<matplotlib.axes._subplots.AxesSubplot object at 0x133a7ded0>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x131edc650>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x1269d6950>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x12d4b99d0>],
       [<matplotlib.axes._subplots.AxesSubplot object at 0x12d89c950>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x139d6b750>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x1240516d0>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x1319390d0>],
       [<matplotlib.axes._subplots.AxesSubplot object at 0x131910050>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x126206750>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x133898110>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x132ca9e90>],
       [<matplotlib.axes._subplots.AxesSubplot object at 0x1344ae750>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x138b37d90>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x1339e6150>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x132811150>]], dtype=object)
_images/scatter_matrix_kde.png

Density Plot

New in version 0.8.0.

You can create density plots using the Series.plot.kde() and DataFrame.plot.kde() methods.

In [83]: ser = pd.Series(np.random.randn(1000))

In [84]: ser.plot.kde()
Out[84]: <matplotlib.axes._subplots.AxesSubplot at 0x124479d10>
_images/kde_plot.png

Andrews Curves

Andrews curves allow one to plot multivariate data as a large number of curves that are created using the attributes of samples as coefficients for Fourier series. By coloring these curves differently for each class it is possible to visualize data clustering. Curves belonging to samples of the same class will usually be closer together and form larger structures.

Note: The “Iris” dataset is available here.

In [85]: from pandas.tools.plotting import andrews_curves

In [86]: data = pd.read_csv('data/iris.data')

In [87]: plt.figure()
Out[87]: <matplotlib.figure.Figure at 0x1246d7950>

In [88]: andrews_curves(data, 'Name')
Out[88]: <matplotlib.axes._subplots.AxesSubplot at 0x130bd6090>
_images/andrews_curves.png

Parallel Coordinates

Parallel coordinates is a plotting technique for plotting multivariate data. It allows one to see clusters in data and to estimate other statistics visually. Using parallel coordinates points are represented as connected line segments. Each vertical line represents one attribute. One set of connected line segments represents one data point. Points that tend to cluster will appear closer together.

In [89]: from pandas.tools.plotting import parallel_coordinates

In [90]: data = pd.read_csv('data/iris.data')

In [91]: plt.figure()
Out[91]: <matplotlib.figure.Figure at 0x131930250>

In [92]: parallel_coordinates(data, 'Name')
Out[92]: <matplotlib.axes._subplots.AxesSubplot at 0x12f006950>
_images/parallel_coordinates.png

Lag Plot

Lag plots are used to check if a data set or time series is random. Random data should not exhibit any structure in the lag plot. Non-random structure implies that the underlying data are not random.

In [93]: from pandas.tools.plotting import lag_plot

In [94]: plt.figure()
Out[94]: <matplotlib.figure.Figure at 0x133c3eb50>

In [95]: data = pd.Series(0.1 * np.random.rand(1000) +
   ....:     0.9 * np.sin(np.linspace(-99 * np.pi, 99 * np.pi, num=1000)))
   ....: 

In [96]: lag_plot(data)
Out[96]: <matplotlib.axes._subplots.AxesSubplot at 0x12edc4a50>
_images/lag_plot.png

Autocorrelation Plot

Autocorrelation plots are often used for checking randomness in time series. This is done by computing autocorrelations for data values at varying time lags. If time series is random, such autocorrelations should be near zero for any and all time-lag separations. If time series is non-random then one or more of the autocorrelations will be significantly non-zero. The horizontal lines displayed in the plot correspond to 95% and 99% confidence bands. The dashed line is 99% confidence band.

In [97]: from pandas.tools.plotting import autocorrelation_plot

In [98]: plt.figure()
Out[98]: <matplotlib.figure.Figure at 0x12eddc290>

In [99]: data = pd.Series(0.7 * np.random.rand(1000) +
   ....:    0.3 * np.sin(np.linspace(-9 * np.pi, 9 * np.pi, num=1000)))
   ....: 

In [100]: autocorrelation_plot(data)
Out[100]: <matplotlib.axes._subplots.AxesSubplot at 0x12eddccd0>
_images/autocorrelation_plot.png

Bootstrap Plot

Bootstrap plots are used to visually assess the uncertainty of a statistic, such as mean, median, midrange, etc. A random subset of a specified size is selected from a data set, the statistic in question is computed for this subset and the process is repeated a specified number of times. Resulting plots and histograms are what constitutes the bootstrap plot.

In [101]: from pandas.tools.plotting import bootstrap_plot

In [102]: data = pd.Series(np.random.rand(1000))

In [103]: bootstrap_plot(data, size=50, samples=500, color='grey')
Out[103]: <matplotlib.figure.Figure at 0x12edc8990>
_images/bootstrap_plot.png

RadViz

RadViz is a way of visualizing multi-variate data. It is based on a simple spring tension minimization algorithm. Basically you set up a bunch of points in a plane. In our case they are equally spaced on a unit circle. Each point represents a single attribute. You then pretend that each sample in the data set is attached to each of these points by a spring, the stiffness of which is proportional to the numerical value of that attribute (they are normalized to unit interval). The point in the plane, where our sample settles to (where the forces acting on our sample are at an equilibrium) is where a dot representing our sample will be drawn. Depending on which class that sample belongs it will be colored differently.

Note: The “Iris” dataset is available here.

In [104]: from pandas.tools.plotting import radviz

In [105]: data = pd.read_csv('data/iris.data')

In [106]: plt.figure()
Out[106]: <matplotlib.figure.Figure at 0x12b0ee1d0>

In [107]: radviz(data, 'Name')
Out[107]: <matplotlib.axes._subplots.AxesSubplot at 0x12edcee10>
_images/radviz.png

Plot Formatting

Most plotting methods have a set of keyword arguments that control the layout and formatting of the returned plot:

In [108]: plt.figure(); ts.plot(style='k--', label='Series');
_images/series_plot_basic2.png

For each kind of plot (e.g. line, bar, scatter) any additional arguments keywords are passed along to the corresponding matplotlib function (ax.plot(), ax.bar(), ax.scatter()). These can be used to control additional styling, beyond what pandas provides.

Controlling the Legend

You may set the legend argument to False to hide the legend, which is shown by default.

In [109]: df = pd.DataFrame(np.random.randn(1000, 4), index=ts.index, columns=list('ABCD'))

In [110]: df = df.cumsum()

In [111]: df.plot(legend=False)
Out[111]: <matplotlib.axes._subplots.AxesSubplot at 0x138cd1950>
_images/frame_plot_basic_noleg.png

Scales

You may pass logy to get a log-scale Y axis.

In [112]: ts = pd.Series(np.random.randn(1000), index=pd.date_range('1/1/2000', periods=1000))

In [113]: ts = np.exp(ts.cumsum())

In [114]: ts.plot(logy=True)
Out[114]: <matplotlib.axes._subplots.AxesSubplot at 0x126824190>
_images/series_plot_logy.png

See also the logx and loglog keyword arguments.

Plotting on a Secondary Y-axis

To plot data on a secondary y-axis, use the secondary_y keyword:

In [115]: df.A.plot()
Out[115]: <matplotlib.axes._subplots.AxesSubplot at 0x132c14550>

In [116]: df.B.plot(secondary_y=True, style='g')
Out[116]: <matplotlib.axes._subplots.AxesSubplot at 0x139b0db10>
_images/series_plot_secondary_y.png

To plot some columns in a DataFrame, give the column names to the secondary_y keyword:

In [117]: plt.figure()
Out[117]: <matplotlib.figure.Figure at 0x124273d90>

In [118]: ax = df.plot(secondary_y=['A', 'B'])

In [119]: ax.set_ylabel('CD scale')
Out[119]: <matplotlib.text.Text at 0x139905210>

In [120]: ax.right_ax.set_ylabel('AB scale')
Out[120]: <matplotlib.text.Text at 0x139bb18d0>
_images/frame_plot_secondary_y.png

Note that the columns plotted on the secondary y-axis is automatically marked with “(right)” in the legend. To turn off the automatic marking, use the mark_right=False keyword:

In [121]: plt.figure()
Out[121]: <matplotlib.figure.Figure at 0x127b66710>

In [122]: df.plot(secondary_y=['A', 'B'], mark_right=False)
Out[122]: <matplotlib.axes._subplots.AxesSubplot at 0x122628850>
_images/frame_plot_secondary_y_no_right.png

Suppressing Tick Resolution Adjustment

pandas includes automatic tick resolution adjustment for regular frequency time-series data. For limited cases where pandas cannot infer the frequency information (e.g., in an externally created twinx), you can choose to suppress this behavior for alignment purposes.

Here is the default behavior, notice how the x-axis tick labelling is performed:

In [123]: plt.figure()
Out[123]: <matplotlib.figure.Figure at 0x127b66450>

In [124]: df.A.plot()
Out[124]: <matplotlib.axes._subplots.AxesSubplot at 0x1245fc2d0>
_images/ser_plot_suppress.png

Using the x_compat parameter, you can suppress this behavior:

In [125]: plt.figure()
Out[125]: <matplotlib.figure.Figure at 0x138a875d0>

In [126]: df.A.plot(x_compat=True)
Out[126]: <matplotlib.axes._subplots.AxesSubplot at 0x12e920550>
_images/ser_plot_suppress_parm.png

If you have more than one plot that needs to be suppressed, the use method in pandas.plot_params can be used in a with statement:

In [127]: plt.figure()
Out[127]: <matplotlib.figure.Figure at 0x1334518d0>

In [128]: with pd.plot_params.use('x_compat', True):
   .....:     df.A.plot(color='r')
   .....:     df.B.plot(color='g')
   .....:     df.C.plot(color='b')
   .....: 
_images/ser_plot_suppress_context.png

Subplots

Each Series in a DataFrame can be plotted on a different axis with the subplots keyword:

In [129]: df.plot(subplots=True, figsize=(6, 6));
_images/frame_plot_subplots.png

Using Layout and Targeting Multiple Axes

The layout of subplots can be specified by layout keyword. It can accept (rows, columns). The layout keyword can be used in hist and boxplot also. If input is invalid, ValueError will be raised.

The number of axes which can be contained by rows x columns specified by layout must be larger than the number of required subplots. If layout can contain more axes than required, blank axes are not drawn. Similar to a numpy array’s reshape method, you can use -1 for one dimension to automatically calculate the number of rows or columns needed, given the other.

In [130]: df.plot(subplots=True, layout=(2, 3), figsize=(6, 6), sharex=False);
_images/frame_plot_subplots_layout.png

The above example is identical to using

In [131]: df.plot(subplots=True, layout=(2, -1), figsize=(6, 6), sharex=False);

The required number of columns (3) is inferred from the number of series to plot and the given number of rows (2).

Also, you can pass multiple axes created beforehand as list-like via ax keyword. This allows to use more complicated layout. The passed axes must be the same number as the subplots being drawn.

When multiple axes are passed via ax keyword, layout, sharex and sharey keywords don’t affect to the output. You should explicitly pass sharex=False and sharey=False, otherwise you will see a warning.

In [132]: fig, axes = plt.subplots(4, 4, figsize=(6, 6));

In [133]: plt.subplots_adjust(wspace=0.5, hspace=0.5);

In [134]: target1 = [axes[0][0], axes[1][1], axes[2][2], axes[3][3]]

In [135]: target2 = [axes[3][0], axes[2][1], axes[1][2], axes[0][3]]

In [136]: df.plot(subplots=True, ax=target1, legend=False, sharex=False, sharey=False);

In [137]: (-df).plot(subplots=True, ax=target2, legend=False, sharex=False, sharey=False);
_images/frame_plot_subplots_multi_ax.png

Another option is passing an ax argument to Series.plot() to plot on a particular axis:

In [138]: fig, axes = plt.subplots(nrows=2, ncols=2)

In [139]: df['A'].plot(ax=axes[0,0]); axes[0,0].set_title('A');

In [140]: df['B'].plot(ax=axes[0,1]); axes[0,1].set_title('B');

In [141]: df['C'].plot(ax=axes[1,0]); axes[1,0].set_title('C');

In [142]: df['D'].plot(ax=axes[1,1]); axes[1,1].set_title('D');
_images/series_plot_multi.png

Plotting With Error Bars

New in version 0.14.

Plotting with error bars is now supported in the DataFrame.plot() and Series.plot()

Horizontal and vertical errorbars can be supplied to the xerr and yerr keyword arguments to plot(). The error values can be specified using a variety of formats.

  • As a DataFrame or dict of errors with column names matching the columns attribute of the plotting DataFrame or matching the name attribute of the Series
  • As a str indicating which of the columns of plotting DataFrame contain the error values
  • As raw values (list, tuple, or np.ndarray). Must be the same length as the plotting DataFrame/Series

Asymmetrical error bars are also supported, however raw error values must be provided in this case. For a M length Series, a Mx2 array should be provided indicating lower and upper (or left and right) errors. For a MxN DataFrame, asymmetrical errors should be in a Mx2xN array.

Here is an example of one way to easily plot group means with standard deviations from the raw data.

# Generate the data
In [143]: ix3 = pd.MultiIndex.from_arrays([['a', 'a', 'a', 'a', 'b', 'b', 'b', 'b'], ['foo', 'foo', 'bar', 'bar', 'foo', 'foo', 'bar', 'bar']], names=['letter', 'word'])

In [144]: df3 = pd.DataFrame({'data1': [3, 2, 4, 3, 2, 4, 3, 2], 'data2': [6, 5, 7, 5, 4, 5, 6, 5]}, index=ix3)

# Group by index labels and take the means and standard deviations for each group
In [145]: gp3 = df3.groupby(level=('letter', 'word'))

In [146]: means = gp3.mean()

In [147]: errors = gp3.std()

In [148]: means
Out[148]: 
             data1  data2
letter word              
a      bar     3.5    6.0
       foo     2.5    5.5
b      bar     2.5    5.5
       foo     3.0    4.5

In [149]: errors
Out[149]: 
                data1     data2
letter word                    
a      bar   0.707107  1.414214
       foo   0.707107  0.707107
b      bar   0.707107  0.707107
       foo   1.414214  0.707107

# Plot
In [150]: fig, ax = plt.subplots()

In [151]: means.plot.bar(yerr=errors, ax=ax)
Out[151]: <matplotlib.axes._subplots.AxesSubplot at 0x12d6b1f90>
_images/errorbar_example.png

Plotting Tables

New in version 0.14.

Plotting with matplotlib table is now supported in DataFrame.plot() and Series.plot() with a table keyword. The table keyword can accept bool, DataFrame or Series. The simple way to draw a table is to specify table=True. Data will be transposed to meet matplotlib’s default layout.

In [152]: fig, ax = plt.subplots(1, 1)

In [153]: df = pd.DataFrame(np.random.rand(5, 3), columns=['a', 'b', 'c'])

In [154]: ax.get_xaxis().set_visible(False)   # Hide Ticks

In [155]: df.plot(table=True, ax=ax)
Out[155]: <matplotlib.axes._subplots.AxesSubplot at 0x126187a50>
_images/line_plot_table_true.png

Also, you can pass different DataFrame or Series for table keyword. The data will be drawn as displayed in print method (not transposed automatically). If required, it should be transposed manually as below example.

In [156]: fig, ax = plt.subplots(1, 1)

In [157]: ax.get_xaxis().set_visible(False)   # Hide Ticks

In [158]: df.plot(table=np.round(df.T, 2), ax=ax)
Out[158]: <matplotlib.axes._subplots.AxesSubplot at 0x130b35150>
_images/line_plot_table_data.png

Finally, there is a helper function pandas.tools.plotting.table to create a table from DataFrame and Series, and add it to an matplotlib.Axes. This function can accept keywords which matplotlib table has.

In [159]: from pandas.tools.plotting import table

In [160]: fig, ax = plt.subplots(1, 1)

In [161]: table(ax, np.round(df.describe(), 2),
   .....:       loc='upper right', colWidths=[0.2, 0.2, 0.2])
   .....: 
Out[161]: <matplotlib.table.Table at 0x128706c90>

In [162]: df.plot(ax=ax, ylim=(0, 2), legend=None)
Out[162]: <matplotlib.axes._subplots.AxesSubplot at 0x127d9c3d0>
_images/line_plot_table_describe.png

Note: You can get table instances on the axes using axes.tables property for further decorations. See the matplotlib table documentation for more.

Colormaps

A potential issue when plotting a large number of columns is that it can be difficult to distinguish some series due to repetition in the default colors. To remedy this, DataFrame plotting supports the use of the colormap= argument, which accepts either a Matplotlib colormap or a string that is a name of a colormap registered with Matplotlib. A visualization of the default matplotlib colormaps is available here.

As matplotlib does not directly support colormaps for line-based plots, the colors are selected based on an even spacing determined by the number of columns in the DataFrame. There is no consideration made for background color, so some colormaps will produce lines that are not easily visible.

To use the cubehelix colormap, we can simply pass 'cubehelix' to colormap=

In [163]: df = pd.DataFrame(np.random.randn(1000, 10), index=ts.index)

In [164]: df = df.cumsum()

In [165]: plt.figure()
Out[165]: <matplotlib.figure.Figure at 0x12ad63250>

In [166]: df.plot(colormap='cubehelix')
Out[166]: <matplotlib.axes._subplots.AxesSubplot at 0x128190490>
_images/cubehelix.png

or we can pass the colormap itself

In [167]: from matplotlib import cm

In [168]: plt.figure()
Out[168]: <matplotlib.figure.Figure at 0x128190250>

In [169]: df.plot(colormap=cm.cubehelix)
Out[169]: <matplotlib.axes._subplots.AxesSubplot at 0x1336d3210>
_images/cubehelix_cm.png

Colormaps can also be used other plot types, like bar charts:

In [170]: dd = pd.DataFrame(np.random.randn(10, 10)).applymap(abs)

In [171]: dd = dd.cumsum()

In [172]: plt.figure()
Out[172]: <matplotlib.figure.Figure at 0x1318edb90>

In [173]: dd.plot.bar(colormap='Greens')
Out[173]: <matplotlib.axes._subplots.AxesSubplot at 0x126255c10>
_images/greens.png

Parallel coordinates charts:

In [174]: plt.figure()
Out[174]: <matplotlib.figure.Figure at 0x128278d90>

In [175]: parallel_coordinates(data, 'Name', colormap='gist_rainbow')
Out[175]: <matplotlib.axes._subplots.AxesSubplot at 0x1287cdcd0>
_images/parallel_gist_rainbow.png

Andrews curves charts:

In [176]: plt.figure()
Out[176]: <matplotlib.figure.Figure at 0x121488890>

In [177]: andrews_curves(data, 'Name', colormap='winter')
Out[177]: <matplotlib.axes._subplots.AxesSubplot at 0x1214a7790>
_images/andrews_curve_winter.png

Plotting directly with matplotlib

In some situations it may still be preferable or necessary to prepare plots directly with matplotlib, for instance when a certain type of plot or customization is not (yet) supported by pandas. Series and DataFrame objects behave like arrays and can therefore be passed directly to matplotlib functions without explicit casts.

pandas also automatically registers formatters and locators that recognize date indices, thereby extending date and time support to practically all plot types available in matplotlib. Although this formatting does not provide the same level of refinement you would get when plotting via pandas, it can be faster when plotting a large number of points.

Note

The speed up for large data sets only applies to pandas 0.14.0 and later.

In [178]: price = pd.Series(np.random.randn(150).cumsum(),
   .....:                   index=pd.date_range('2000-1-1', periods=150, freq='B'))
   .....: 

In [179]: ma = price.rolling(20).mean()

In [180]: mstd = price.rolling(20).std()

In [181]: plt.figure()
Out[181]: <matplotlib.figure.Figure at 0x12817cd50>

In [182]: plt.plot(price.index, price, 'k')
Out[182]: [<matplotlib.lines.Line2D at 0x12147b790>]

In [183]: plt.plot(ma.index, ma, 'b')
Out[183]: [<matplotlib.lines.Line2D at 0x121494610>]

In [184]: plt.fill_between(mstd.index, ma-2*mstd, ma+2*mstd, color='b', alpha=0.2)
Out[184]: <matplotlib.collections.PolyCollection at 0x121494f10>
_images/bollinger.png

Trellis plotting interface

Warning

The rplot trellis plotting interface is deprecated and will be removed in a future version. We refer to external packages like seaborn for similar but more refined functionality.

The docs below include some example on how to convert your existing code to seaborn.

Note

The tips data set can be downloaded here. Once you download it execute

tips_data = pd.read_csv('tips.csv')

from the directory where you downloaded the file.

We import the rplot API:

In [185]: import pandas.tools.rplot as rplot

Examples

RPlot was an API for producing Trellis plots. These plots allow you to arrange data in a rectangular grid by values of certain attributes. In the example below, data from the tips data set is arranged by the attributes ‘sex’ and ‘smoker’. Since both of those attributes can take on one of two values, the resulting grid has two columns and two rows. A histogram is displayed for each cell of the grid.

In [186]: plt.figure()
Out[186]: <matplotlib.figure.Figure at 0x13189f810>

In [187]: plot = rplot.RPlot(tips_data, x='total_bill', y='tip')

In [188]: plot.add(rplot.TrellisGrid(['sex', 'smoker']))

In [189]: plot.add(rplot.GeomHistogram())

In [190]: plot.render(plt.gcf())
Out[190]: <matplotlib.figure.Figure at 0x13189f810>
_images/rplot1_tips.png

A similar plot can be made with seaborn using the FacetGrid object, resulting in the following image:

import seaborn as sns
g = sns.FacetGrid(tips_data, row="sex", col="smoker")
g.map(plt.hist, "total_bill")
_images/rplot-seaborn-example1.png

Example below is the same as previous except the plot is set to kernel density estimation. A seaborn example is included beneath.

In [191]: plt.figure()
Out[191]: <matplotlib.figure.Figure at 0x126a68910>

In [192]: plot = rplot.RPlot(tips_data, x='total_bill', y='tip')

In [193]: plot.add(rplot.TrellisGrid(['sex', 'smoker']))

In [194]: plot.add(rplot.GeomDensity())

In [195]: plot.render(plt.gcf())
Out[195]: <matplotlib.figure.Figure at 0x126a68910>
_images/rplot2_tips.png
g = sns.FacetGrid(tips_data, row="sex", col="smoker")
g.map(sns.kdeplot, "total_bill")
_images/rplot-seaborn-example2.png

The plot below shows that it is possible to have two or more plots for the same data displayed on the same Trellis grid cell.

In [196]: plt.figure()
Out[196]: <matplotlib.figure.Figure at 0x12a83b490>

In [197]: plot = rplot.RPlot(tips_data, x='total_bill', y='tip')

In [198]: plot.add(rplot.TrellisGrid(['sex', 'smoker']))

In [199]: plot.add(rplot.GeomScatter())

In [200]: plot.add(rplot.GeomPolyFit(degree=2))

In [201]: plot.render(plt.gcf())
Out[201]: <matplotlib.figure.Figure at 0x12a83b490>
_images/rplot3_tips.png

A seaborn equivalent for a simple scatter plot:

g = sns.FacetGrid(tips_data, row="sex", col="smoker")
g.map(plt.scatter, "total_bill", "tip")
_images/rplot-seaborn-example3.png

and with a regression line, using the dedicated seaborn regplot function:

g = sns.FacetGrid(tips_data, row="sex", col="smoker", margin_titles=True)
g.map(sns.regplot, "total_bill", "tip", order=2)
_images/rplot-seaborn-example3b.png

Below is a similar plot but with 2D kernel density estimation plot superimposed, followed by a seaborn equivalent:

In [202]: plt.figure()
Out[202]: <matplotlib.figure.Figure at 0x127d42050>

In [203]: plot = rplot.RPlot(tips_data, x='total_bill', y='tip')

In [204]: plot.add(rplot.TrellisGrid(['sex', 'smoker']))

In [205]: plot.add(rplot.GeomScatter())

In [206]: plot.add(rplot.GeomDensity2D())

In [207]: plot.render(plt.gcf())
Out[207]: <matplotlib.figure.Figure at 0x127d42050>
_images/rplot4_tips.png
g = sns.FacetGrid(tips_data, row="sex", col="smoker")
g.map(plt.scatter, "total_bill", "tip")
g.map(sns.kdeplot, "total_bill", "tip")
_images/rplot-seaborn-example4.png

It is possible to only use one attribute for grouping data. The example above only uses ‘sex’ attribute. If the second grouping attribute is not specified, the plots will be arranged in a column.

In [208]: plt.figure()
Out[208]: <matplotlib.figure.Figure at 0x12f47cb10>

In [209]: plot = rplot.RPlot(tips_data, x='total_bill', y='tip')

In [210]: plot.add(rplot.TrellisGrid(['sex', '.']))

In [211]: plot.add(rplot.GeomHistogram())

In [212]: plot.render(plt.gcf())
Out[212]: <matplotlib.figure.Figure at 0x12f47cb10>
_images/rplot5_tips.png

If the first grouping attribute is not specified the plots will be arranged in a row.

In [213]: plt.figure()
Out[213]: <matplotlib.figure.Figure at 0x12f56a550>

In [214]: plot = rplot.RPlot(tips_data, x='total_bill', y='tip')

In [215]: plot.add(rplot.TrellisGrid(['.', 'smoker']))

In [216]: plot.add(rplot.GeomHistogram())

In [217]: plot.render(plt.gcf())
Out[217]: <matplotlib.figure.Figure at 0x12f56a550>
_images/rplot6_tips.png

In seaborn, this can also be done by only specifying one of the row and col arguments.

In the example below the colour and shape of the scatter plot graphical objects is mapped to ‘day’ and ‘size’ attributes respectively. You use scale objects to specify these mappings. The list of scale classes is given below with initialization arguments for quick reference.

In [218]: plt.figure()
Out[218]: <matplotlib.figure.Figure at 0x1300d9590>

In [219]: plot = rplot.RPlot(tips_data, x='tip', y='total_bill')

In [220]: plot.add(rplot.TrellisGrid(['sex', 'smoker']))

In [221]: plot.add(rplot.GeomPoint(size=80.0, colour=rplot.ScaleRandomColour('day'), shape=rplot.ScaleShape('size'), alpha=1.0))

In [222]: plot.render(plt.gcf())
Out[222]: <matplotlib.figure.Figure at 0x1300d9590>
_images/rplot7_tips.png

This can also be done in seaborn, at least for 3 variables:

g = sns.FacetGrid(tips_data, row="sex", col="smoker", hue="day")
g.map(plt.scatter, "tip", "total_bill")
g.add_legend()
_images/rplot-seaborn-example6.png