Onnx运行时错误“ SVC转换器不支持与前n个整数不同的类”是什么意思？

如何解决Onnx运行时错误“ SVC转换器不支持与前n个整数不同的类”是什么意思？

我正在尝试将sklearn SVC模型转换/存储为.onnx文件，并且遇到了我不理解的运行时错误。通过sklearn随机森林分类器和sklearn k-NN分类器，我已经能够有效地使用相同的代码而不会出现错误。我收到一个我不理解的奇怪的onnx运行时错误。感谢您提供有关此错误的任何帮助。

我首先在下面发布了运行文件svm_time.py的输出，然后在下面显示了svm_time.py文件中包含的代码。

谢谢。

df %>%
   mutate(x2 = case_when(a == 3 ~  0.05 + x))

python3 svm_time.py 
'train_model'  4809.58 ms
train score is:  0.8765468473777254
val Accuracy is:  0.7037037037037037
Traceback (most recent call last):
  File "svm_time.py",line 97,in <module>
    main()
  File "svm_time.py",line 91,in main
    onx = convert_sklearn(clf,initial_types=initial_type)
  File "/home/matt/anaconda3/envs/venv/lib/python3.7/site-packages/skl2onnx/convert.py",line 154,in convert_sklearn
    dtype=dtype,options=options)
  File "/home/matt/anaconda3/envs/venv/lib/python3.7/site-packages/skl2onnx/common/_topology.py",line 1054,in convert_topology
    conv(scope,operator,container)
  File "/home/matt/anaconda3/envs/venv/lib/python3.7/site-packages/skl2onnx/common/_registration.py",line 29,in __call__
    return self._fct(*args)
  File "/home/matt/anaconda3/envs/venv/lib/python3.7/site-packages/skl2onnx/operator_converters/support_vector_machines.py",line 221,in convert_sklearn_svm_classifier
    "Classes different from first n integers are not supported "
RuntimeError: Classes different from first n integers are not supported in SVC converter.

解决方法

在我看来，这可能是Onnx和sklearn的兼容性问题。

1）。 https://github.com/onnx/sklearn-onnx/issues/302 2）。 https://github.com/onnx/sklearn-onnx/blob/master/skl2onnx/operator_converters/support_vector_machines.py#L17

基于这两个来源，我更改了代码，以包括OVO决策函数形状而不是OVR，现在，至少在运行svm_time.py文件时，将保存一个.onnx文件。

import numpy as np
import pandas as pd
import plotly.graph_objects as go
import plotly.express as px
import datetime

pd.set_option('display.max_rows',None)

# data sample
nperiods = 200
np.random.seed(123)
df = pd.DataFrame(np.random.randint(-10,12,size=(nperiods,4)),columns=list('ABCD'))
datelist = pd.date_range(datetime.datetime(2020,1,1).strftime('%Y-%m-%d'),periods=nperiods).tolist()
df['dates'] = datelist 
df = df.set_index(['dates'])
df.index = pd.to_datetime(df.index)
df.iloc[0] = 0
df = df.cumsum().reset_index()

# function to set background color for a
# specified variable and a specified level



# plotly setup
fig = px.line(df,x='dates',y=df.columns[1:])
fig.update_xaxes(showgrid=True,gridwidth=1,gridcolor='rgba(0,255,0.1)')
fig.update_yaxes(showgrid=True,0.1)')

def bgLevels(fig,variable,level,mode,fillcolor,layer):
    """
    Set a specified color as background for given
    levels of a specified variable using a shape.
    
    Keyword arguments:
    ==================
    fig -- plotly figure
    variable -- column name in a pandas dataframe
    level -- int or float
    mode -- set threshold above or below
    fillcolor -- any color type that plotly can handle
    layer -- position of shape in plotly fiugre,like "below"
    
    """
    
    if mode == 'above':
        m = df[variable].gt(level)
    
    if mode == 'below':
        m = df[variable].lt(level)
        
    df1 = df[m].groupby((~m).cumsum())['dates'].agg(['first','last'])

    for index,row in df1.iterrows():
        #print(row['first'],row['last'])
        fig.add_shape(type="rect",xref="x",yref="paper",x0=row['first'],y0=0,x1=row['last'],y1=1,line=dict(color="rgba(0,0)",width=3,),fillcolor=fillcolor,layer=layer) 
    return(fig)

fig = bgLevels(fig = fig,variable = 'A',level = 100,mode = 'above',fillcolor = 'rgba(100,100,0.2)',layer = 'below')

fig = bgLevels(fig = fig,variable = 'D',level = -60,mode = 'below',fillcolor = 'rgba(255,layer = 'below')

fig.show()