Python seaborn.catplot() Examples

def _plot_results_per_citation_task(results_df, save_cfg):
    """Plot scatter plot of accuracy for each condition and task.
    """
    fig, ax = plt.subplots(figsize=(save_cfg['text_width'], 
                                    save_cfg['text_height'] * 1.3))
    # figsize = plt.rcParams.get('figure.figsize')
    # fig, ax = plt.subplots(figsize=(figsize[0], figsize[1] * 4))
    # Need to make the graph taller otherwise the y axis labels are on top of
    # each other.
    sns.catplot(y='citation_task', x='Result', hue='model_type', data=results_df, 
                ax=ax)
    ax.set_xlabel('accuracy')
    ax.set_ylabel('')
    plt.tight_layout()

    if save_cfg is not None:
        savename = 'reported_results'
        fname = os.path.join(save_cfg['savepath'], savename)
        fig.savefig(fname + '.' + save_cfg['format'], **save_cfg)

    return ax 

def _plot_results_accuracy_diff_scatter(results_df, save_cfg):
    """Plot difference in accuracy for each condition/task as a scatter plot.
    """
    fig, ax = plt.subplots(figsize=(save_cfg['text_width'], 
                                    save_cfg['text_height'] * 1.3))
    # figsize = plt.rcParams.get('figure.figsize')
    # fig, ax = plt.subplots(figsize=(figsize[0], figsize[1] * 2))
    sns.catplot(y='Task', x='acc_diff', data=results_df, ax=ax)
    ax.set_xlabel('Accuracy difference')
    ax.set_ylabel('')
    ax.axvline(0, c='k', alpha=0.2)
    plt.tight_layout()

    if save_cfg is not None:
        savename = 'reported_accuracy_diff_scatter'
        fname = os.path.join(save_cfg['savepath'], savename)
        fig.savefig(fname + '.' + save_cfg['format'], **save_cfg)

    return ax 

def plot_hardware(df, save_cfg=cfg.saving_config):
    """Plot bar graph showing the hardware used in the study.
    """
    col = 'EEG Hardware'
    hardware_df = ut.split_column_with_multiple_entries(
        df, col, ref_col='Citation', sep=',', lower=False)

    # Remove N/Ms because they make it hard to see anything
    hardware_df = hardware_df[hardware_df[col] != 'N/M']
    
    # Add low cost column
    hardware_df['Low-cost'] = False
    low_cost_devices = ['EPOC (Emotiv)', 'OpenBCI (OpenBCI)', 'Muse (InteraXon)', 
                        'Mindwave Mobile (Neurosky)', 'Mindset (NeuroSky)']
    hardware_df.loc[hardware_df[col].isin(low_cost_devices), 
                    'Low-cost'] = True

    fig, ax = plt.subplots(figsize=(save_cfg['text_width'] / 4 * 2, 
                                    save_cfg['text_height'] / 5 * 2))
    sns.countplot(hue=hardware_df['Low-cost'], y=hardware_df[col], ax=ax,
                  order=hardware_df[col].value_counts().index, 
                  dodge=False)
    # sns.catplot(row=hardware_df['low_cost'], y=hardware_df['hardware'])
    ax.set_xlabel('Number of papers')
    ax.set_ylabel('')
    plt.tight_layout()

    if save_cfg is not None:
        fname = os.path.join(save_cfg['savepath'], 'hardware')
        fig.savefig(fname + '.' + save_cfg['format'], **save_cfg)

    return ax 

def plot_discovered_missed_clusters(dataframe, file_path=''):
  """Plot location clusters comparing agents missed and discovered incidents."""
  plot_height = 5
  aspect_ratio = 1.3
  sns.catplot(
      x='param_value',
      y='missed_incidents',
      data=dataframe,
      hue='agent_type',
      palette='deep',
      height=plot_height,
      aspect=aspect_ratio,
      s=8,
      legend=False)
  plt.xlabel('Dynamic factor', fontsize=LARGE_FONTSIZE)
  plt.ylabel('Missed incidents for each location', fontsize=LARGE_FONTSIZE)
  plt.xticks(fontsize=MEDIUM_FONTSIZE)
  plt.yticks(fontsize=MEDIUM_FONTSIZE)
  plt.legend(fontsize=MEDIUM_FONTSIZE, title_fontsize=MEDIUM_FONTSIZE)
  plt.tight_layout()
  plt.savefig(file_path + '_missed.pdf')

  sns.catplot(
      x='param_value',
      y='discovered_incidents',
      data=dataframe,
      hue='agent_type',
      height=plot_height,
      aspect=aspect_ratio,
      s=8,
      legend=False)
  plt.xlabel('Dynamic factor', fontsize=LARGE_FONTSIZE)
  plt.ylabel('Discovered incidents for each location', fontsize=LARGE_FONTSIZE)
  plt.xticks(fontsize=MEDIUM_FONTSIZE)
  plt.yticks(fontsize=MEDIUM_FONTSIZE)
  plt.legend(fontsize=MEDIUM_FONTSIZE, title_fontsize=MEDIUM_FONTSIZE)
  plt.tight_layout()
  plt.savefig(file_path + '_discovered.pdf') 

def plot_total_miss_discovered(dataframe, file_path=''):
  """Plot bar charts comparing agents total missed and discovered incidents."""
  plot_height = 5
  aspect_ratio = 1.3
  sns.set_style('whitegrid')
  sns.despine()

  sns.catplot(
      x='param_value',
      y='total_missed',
      data=dataframe,
      hue='agent_type',
      kind='bar',
      palette='muted',
      height=plot_height,
      aspect=aspect_ratio,
      legend=False)
  plt.xlabel('Dynamic factor', fontsize=LARGE_FONTSIZE)
  plt.ylabel('Total missed incidents', fontsize=LARGE_FONTSIZE)
  plt.xticks(fontsize=MEDIUM_FONTSIZE)
  plt.yticks(fontsize=MEDIUM_FONTSIZE)
  plt.legend(fontsize=MEDIUM_FONTSIZE, title_fontsize=MEDIUM_FONTSIZE)
  plt.savefig(file_path + '_missed.pdf', bbox_inches='tight')

  sns.catplot(
      x='param_value',
      y='total_discovered',
      data=dataframe,
      hue='agent_type',
      kind='bar',
      palette='muted',
      height=plot_height,
      aspect=aspect_ratio,
      legend=False)
  plt.xlabel('Dynamic factor', fontsize=LARGE_FONTSIZE)
  plt.ylabel('Total discovered incidents', fontsize=LARGE_FONTSIZE)
  plt.xticks(fontsize=MEDIUM_FONTSIZE)
  plt.yticks(fontsize=MEDIUM_FONTSIZE)
  plt.legend(fontsize=MEDIUM_FONTSIZE, title_fontsize=MEDIUM_FONTSIZE)
  plt.savefig(file_path + '_discovered.pdf', bbox_inches='tight') 

def plot_discovered_occurred_ratio_range(dataframe, file_path=''):
  """Plot the range of discovered incidents/occurred range between locations."""
  plot_height = 5
  aspect_ratio = 1.3
  sns.set_style('whitegrid')
  sns.despine()

  sns.catplot(
      x='param_value',
      y='discovered/occurred range',
      data=dataframe,
      hue='agent_type',
      kind='bar',
      palette='muted',
      height=plot_height,
      aspect=aspect_ratio)
  plt.xlabel('Dynamic factor', fontsize=LARGE_FONTSIZE)
  plt.ylabel('Discovered/occurred range', fontsize=LARGE_FONTSIZE)
  plt.xticks(fontsize=MEDIUM_FONTSIZE)
  plt.yticks(fontsize=MEDIUM_FONTSIZE)
  # plt.legend(fontsize=MEDIUM_FONTSIZE, title_fontsize=MEDIUM_FONTSIZE)
  plt.savefig(file_path + '.pdf', bbox_inches='tight')

  sns.catplot(
      x='param_value',
      y='discovered/occurred range weighted',
      data=dataframe,
      hue='agent_type',
      kind='bar',
      palette='muted',
      height=plot_height,
      aspect=aspect_ratio)
  plt.xlabel('Dynamic factor', fontsize=LARGE_FONTSIZE)
  plt.ylabel('Delta', fontsize=LARGE_FONTSIZE)
  plt.xticks(fontsize=MEDIUM_FONTSIZE)
  plt.yticks(fontsize=MEDIUM_FONTSIZE)
  # plt.legend(fontsize=MEDIUM_FONTSIZE, title_fontsize=MEDIUM_FONTSIZE)
  plt.savefig(file_path + '_weighted.pdf', bbox_inches='tight') 

def plot_discovered_occurred_ratio_locations(dataframe, file_path=''):
  """Plot the discovered incidents/occurred ratio for each location."""
  plot_height = 5
  aspect_ratio = 1.3
  sns.despine()

  sns.set(style='ticks')
  sns.catplot(
      x='param_value',
      y='discovered/occurred',
      data=dataframe,
      hue='agent_type',
      height=plot_height,
      aspect=aspect_ratio,
      s=8)
  plt.xlabel('Dynamic factor', fontsize=LARGE_FONTSIZE)
  plt.ylabel('Discovered/occurred', fontsize=LARGE_FONTSIZE)
  plt.xticks(fontsize=MEDIUM_FONTSIZE)
  plt.yticks(fontsize=MEDIUM_FONTSIZE)
  # plt.legend(fontsize=MEDIUM_FONTSIZE, title_fontsize=MEDIUM_FONTSIZE)
  plt.savefig(file_path + '.pdf', bbox_inches='tight')

  sns.set(style='ticks')
  sns.catplot(
      x='param_value',
      y='discovered/occurred weighted',
      data=dataframe,
      hue='agent_type',
      height=plot_height,
      aspect=aspect_ratio,
      s=8)
  plt.xlabel('Dynamic factor', fontsize=LARGE_FONTSIZE)
  plt.ylabel('Discovered/occurred weighted', fontsize=LARGE_FONTSIZE)
  plt.xticks(fontsize=MEDIUM_FONTSIZE)
  plt.yticks(fontsize=MEDIUM_FONTSIZE)
  # plt.legend(fontsize=MEDIUM_FONTSIZE, title_fontsize=MEDIUM_FONTSIZE)
  plt.savefig(file_path + '_weighted.pdf', bbox_inches='tight') 

def analyze_pcap(rl_algos, tcp_algos, plt_name, runs, data_dir):
    import matplotlib
    matplotlib.use('Agg')
    import matplotlib.pyplot as plt
    import seaborn as sns

    algos = rl_algos + tcp_algos
    host_rtt = {}
    for algo in algos:
        host_rtt[algo] = process_rtt_files(data_dir, runs, algo)
    pcap_df = pd.DataFrame.from_dict(host_rtt, orient='index')
    pcap_df = pd.melt(pcap_df.reset_index(), id_vars='index',
                      var_name="Metric",
                      value_name="Average RTT (ms)")
    pcap_df = pcap_df.rename(columns={'index': 'Algorithm'})
    # Convert to milliseconds
    # pcap_df = pcap_df.div(1e6)
    fig = sns.catplot(x='Metric', y='Average RTT (ms)',
                      hue="Algorithm", data=pcap_df, kind='bar')
    from itertools import cycle
    hatches = cycle(["/", "-", "+", "x", '-', '+', 'x', 'O', '.'])

    num_locations = len(pcap_df.Metric.unique())
    for i, patch in enumerate(fig.ax.patches):
        # Blue bars first, then green bars
        if i % num_locations == 0:
            hatch = next(hatches)
        patch.set_hatch(hatch)
    plt_name += "_rtt"
    log.info("Saving plot %s" % plt_name)
    plt.savefig(plt_name + ".png", bbox_inches='tight', pad_inches=0.05)
    plt.savefig(plt_name + ".pdf", bbox_inches='tight', pad_inches=0.05)
    plt.gcf().clear() 

def _plot_results_accuracy_per_domain(results_df, diff_df, save_cfg):
    """Make scatterplot + boxplot to show accuracy difference by domain.
    """
    fig, axes = plt.subplots(
        nrows=2, ncols=1, sharex=True, 
        figsize=(save_cfg['text_width'], save_cfg['text_height'] / 3), 
        gridspec_kw = {'height_ratios':[5, 1]})

    results_df['Main domain'] = results_df['Main domain'].apply(
        ut.wrap_text, max_char=20)

    sns.catplot(y='Main domain', x='acc_diff', s=3, jitter=True, 
                data=results_df, ax=axes[0])
    axes[0].set_xlabel('')
    axes[0].set_ylabel('')
    axes[0].axvline(0, c='k', alpha=0.2)

    sns.boxplot(x='acc_diff', data=diff_df, ax=axes[1])
    sns.swarmplot(x='acc_diff', data=diff_df, color="0", size=2, ax=axes[1])
    axes[1].axvline(0, c='k', alpha=0.2)
    axes[1].set_xlabel('Accuracy difference')

    fig.subplots_adjust(wspace=0, hspace=0.02)
    plt.tight_layout()

    logger.info('Number of studies included in the accuracy improvement analysis: {}'.format(
        results_df.shape[0]))
    median = diff_df['acc_diff'].median()
    iqr = diff_df['acc_diff'].quantile(.75) - diff_df['acc_diff'].quantile(.25)
    logger.info('Median gain in accuracy: {:.6f}'.format(median))
    logger.info('Interquartile range of the gain in accuracy: {:.6f}'.format(iqr))
    best_improvement = diff_df.nlargest(3, 'acc_diff')
    logger.info('Best improvement in accuracy: {}, in {}'.format(
        best_improvement['acc_diff'].values[0], 
        best_improvement['Citation'].values[0]))
    logger.info('Second best improvement in accuracy: {}, in {}'.format(
        best_improvement['acc_diff'].values[1], 
        best_improvement['Citation'].values[1]))
    logger.info('Third best improvement in accuracy: {}, in {}'.format(
        best_improvement['acc_diff'].values[2], 
        best_improvement['Citation'].values[2]))

    if save_cfg is not None:
        savename = 'reported_accuracy_per_domain'
        fname = os.path.join(save_cfg['savepath'], savename)
        fig.savefig(fname + '.' + save_cfg['format'], **save_cfg)

    return axes 

def visualisationDF(df):    
    dataFrameInfoPrint(df)
    #graph-01
    # df['shapelyArea'].plot.hist(alpha=0.5)    
    #graph-02
    # df['shapelyArea'].plot.kde()    
    #graph-03
    # df[['shapelyLength','shapeIdx']].plot.scatter('shapelyLength','shapeIdx')    
    #normalize data in a range of columns
    cols_to_norm=['shapeIdx', 'FRAC']
    df[cols_to_norm]=df[cols_to_norm].apply(lambda x: (x - x.min()) / (x.max() - x.min()))
    
    a='shapeIdx'
    b='FRAC'
    c='park_class'
    
    #graph-04
    # sns.jointplot(a,b,df,kind='hex')
    
    #graph-05
    # sns.jointplot(a, b, df, kind='kde')
    
    #graph-06
    # sns.catplot(x='park_class',y=a,data=df)
    
    #graph-07
    '''
    # Initialize the figure
    f, ax = plt.subplots()
    sns.despine(bottom=True, left=True)
    # Show each observation with a scatterplot
    sns.stripplot(x=a, y=c, hue=c,data=df, dodge=True, alpha=.25, zorder=1)    
    # Show the conditional means
    sns.pointplot(x=a, y=c, hue=c,data=df, dodge=.532, join=False, palette="dark",markers="d", scale=.75, ci=None)
    # Improve the legend 
    handles, labels = ax.get_legend_handles_labels()
    ax.legend(handles[3:], labels[3:], title=b,handletextpad=0, columnspacing=1,loc="lower right", ncol=3, frameon=True)
    '''
    
    #graph-08
    # sns.catplot(x=c,y=a,data=df,kind='box')
    
    #graph-09
    # sns.catplot(x=c,y=a,data=df,kind='violin')
    
    #graph-10
    '''
    f, axs = plt.subplots(1, 2, figsize=(12, 6))
    # First axis    
    df[b].plot.hist(ax=axs[0])
    # Second axis
    df[b].plot.kde(ax=axs[1])
    # Title
    f.suptitle(b)
    # Display
    plt.show()
    '''        

#从新定义栅格投影，参考投影为vector .shp文件