带光滑线的Python散点图

Question

我有如下所示的数据(经过大量预处理后获得)

请找数据

d = {'token': {361: '180816_031', 119: '180816_031', 101: '180816_031', 135: '180816_031', 292: '180816_031',
           133: '180816_031', 99: '180816_031', 270: '180816_031', 19: '180816_031', 382: '180816_031',
           414: '180816_031', 267: '180816_031', 218: '180816_031', 398: '180816_031', 287: '180816_031',
           155: '180816_031', 392: '180816_031', 265: '180816_031', 239: '180816_031', 237: '180816_031'},
 'station': {361: 'deneb', 119: 'callisto', 101: 'callisto', 135: 'callisto', 292: 'callisto', 133: 'deneb',
             99: 'callisto', 270: 'callisto', 19: 'deneb', 382: 'callisto', 414: 'deneb', 267: 'callisto',
             218: 'deneb', 398: 'callisto', 287: 'deneb', 155: 'deneb', 392: 'deneb', 265: 'callisto',
             239: 'callisto', 237: 'callisto'},
 'cycle_number': {361: 'cycle09', 119: 'cycle06', 101: 'cycle04', 135: 'cycle01', 292: 'cycle04', 133: 'cycle05',
                  99: 'cycle06', 270: 'cycle07', 19: 'cycle04', 382: 'cycle08', 414: 'cycle04', 267: 'cycle10',
                  218: 'cycle07', 398: 'cycle08', 287: 'cycle09', 155: 'cycle08', 392: 'cycle06', 265: 'cycle02',
                  239: 'cycle09', 237: 'cycle07'},
 'variable': {361: 'adj_high_quality_reads', 119: 'short_pass', 101: 'short_pass', 135: 'cell_mask_bilayers_sum',
              292: 'adj_active_polymerase', 133: 'cell_mask_bilayers_sum', 99: 'short_pass',
              270: 'adj_active_polymerase', 19: 'Unnamed: 0', 382: 'adj_high_quality_reads',
              414: 'num_align_high_quality_reads', 267: 'adj_active_polymerase', 218: 'adj_single_pores',
              398: 'num_align_high_quality_reads', 287: 'adj_active_polymerase', 155: 'cell_mask_bilayers_sum',
              392: 'num_align_high_quality_reads', 265: 'adj_active_polymerase', 239: 'adj_single_pores',
              237: 'adj_single_pores'},
 'value': {361: 99704.0, 119: 2072785.0, 101: 2061059.0, 135: 1682208.0, 292: 675306.0, 133: 1714292.0,
           99: 2072785.0, 270: 687988.0, 19: 19.0, 382: np.nan, 414: 285176.0, 267: 86914.0, 218: 948971.0,
           398: 405196.0, 287: 137926.0, 155: 1830032.0, 392: 480081.0, 265: 951689.0, 239: 681452.0,
           237: 882671.0}}

数据：

          token   station cycle_number                      variable  \
19   180816_031     deneb      cycle04                    Unnamed: 0   
99   180816_031  callisto      cycle06                    short_pass   
101  180816_031  callisto      cycle04                    short_pass   
119  180816_031  callisto      cycle06                    short_pass   
133  180816_031     deneb      cycle05        cell_mask_bilayers_sum   
135  180816_031  callisto      cycle01        cell_mask_bilayers_sum   
155  180816_031     deneb      cycle08        cell_mask_bilayers_sum   
218  180816_031     deneb      cycle07              adj_single_pores   
237  180816_031  callisto      cycle07              adj_single_pores   
239  180816_031  callisto      cycle09              adj_single_pores   
265  180816_031  callisto      cycle02         adj_active_polymerase   
267  180816_031  callisto      cycle10         adj_active_polymerase   
270  180816_031  callisto      cycle07         adj_active_polymerase   
287  180816_031     deneb      cycle09         adj_active_polymerase   
292  180816_031  callisto      cycle04         adj_active_polymerase   
361  180816_031     deneb      cycle09        adj_high_quality_reads   
382  180816_031  callisto      cycle08        adj_high_quality_reads   
392  180816_031     deneb      cycle06  num_align_high_quality_reads   
398  180816_031  callisto      cycle08  num_align_high_quality_reads   
414  180816_031     deneb      cycle04  num_align_high_quality_reads   

         value  
19        19.0  
99   2072785.0  
101  2061059.0  
119  2072785.0  
133  1714292.0  
135  1682208.0  
155  1830032.0  
218   948971.0  
237   882671.0  
239   681452.0  
265   951689.0  
267    86914.0  
270   687988.0  
287   137926.0  
292   675306.0  
361    99704.0  
382        NaN  
392   480081.0  
398   405196.0  
414   285176.0  

我试图用光滑的线条来创建散点图。

fig,ax = plt.subplots()
fig.set_size_inches(16,4)

#to get different colors for each of the `variable` value assign the variable to hue
g2=sns.lmplot(x='cycle_number',y='value',data=df, hue='variable',  size=4, aspect=5)

此代码只为散点图提供一个值，但我的预期输出如下所示

预期输出：

​

​

尝试结果

尝试1

我试图创建条形图(在一些帮助下)&我是成功的，但是用散点图我做不到。

下面的代码可将其转换为条形图

df1 = df.groupby(['token','variable']).agg({'value': 'mean'})
df1.reset_index(inplace=True)
df1.sort_values('value',inplace=True,ascending=False)
fig,ax = plt.subplots()
fig.set_size_inches(16,8)
#to get different colors for each of the variable assign the variable to hue
g=sns.barplot(x='token',y='value',data=df1, hue='variable',ax=ax)
#Code for to put legend outside the plot
box = ax.get_position()
ax.set_position([box.x0, box.y0, box.width * 0.8, box.height])
# Put a legend to the right of the current axis
ax.legend(loc='center left', bbox_to_anchor=(1, 0.5))
# Adding respective values to the top of each bar
for p in ax.patches: 
    ax.annotate("%d" % p.get_height(), (p.get_x() + p.get_width() / 2, p.get_height()),
                ha='center', va='center', fontsize=11, color='black', xytext=(0, 10), 
                textcoords='offset points',fontweight='bold')
plt.show()

尝试2

g2=sns.lmplot(x='cycle_number',y='value',data=df), this gives error 
ValueError: could not convert string to float: 'cycle10'

我知道错误在这里意味着什么，但是试图复制到输出代码时我感到无能为力。

尝试3：

sns.lmplot('cycle_number', 'value', data=df, hue='variable', fit_reg=False)

生成的输出:空白网格

Answer 1

用途：

sns.pointplot('cycle_number', 'value', data=df, hue='variable')

​

​

documnetation: https://seaborn.pydata.org/generated/seaborn.pointplot.html

使用此输出与预期输出生成输出。

​

​

预期产出

试试这个：

df = pd.DataFrame(d)

df['cycle_number'] = df['cycle_number'].str.replace('cycle', '')
df['cycle_number'] = df['cycle_number'].apply(pd.to_numeric)
print(df)

fig, ax = plt.subplots()
fig.set_size_inches(16, 4)
# sns.pointplot('cycle_number', 'value', data=df, hue='variable', err_style="bars", ci=68)
sns.lmplot('cycle_number', 'value', data=df, hue='variable', ci=None, order=2, truncate=True)
# use order = 5 to see more curve

order=2的输出

​

​

根据最新的共享代码(用于order=2__)输出

​

• 图例与图形区域重叠。

产出4(用于order=5)*：

​

图形曲线非常精细，只是图例与绘图区域重叠。