编码随机森林调查数据的字符串特征

Question

我有一个包含 80 个特征的调查数据集，其中许多是字符串。 喜欢将它们编码为 2 个目标的随机森林，1 个二进制数字分类 (1, 0) 和其他连续的。 LabelEncoder() 将创建我不想要的序数类别。 OneHotEncoder() 会将列吹到一个疯狂的高数。 即使必须为每个目标进行不同编码的方法也很好，我只想跳过这第一步。 谢谢

data.shape
(2946, 80)

data.nunique()
LOAN_AMT                       492
Business Title                1244
Civil Service Title            312
Division/Work Unit             678
Job Description               1608
Minimum Qual Requirements      336
Preferred Skills              1282
Additional Information         681
To Apply                       893
Hours/Shift                    181
Residency Requirement           50
Posting Date                   493
Posting Updated                487
Process Date                     1
FormalEducation                  9
UndergradMajor                  12
CompanySize                      8
DevType                        810
YearsCoding                     11
YearsCodingProf                 11
HopeFiveYears                    7
JobSearchStatus                  3
LastNewJob                       5
UpdateCV                         8
CommunicationTools             348
TimeFullyProductive              6
EducationTypes                 264
SelfTaughtTypes                226
TimeAfterBootcamp                8
HackathonReasons                98
AgreeDisagree1                   5
AgreeDisagree2                   5
AgreeDisagree3                   5
LanguageWorkedWith            1578
LanguageDesireNextYear        1607
DatabaseWorkedWith             691
DatabaseDesireNextYear         819
PlatformWorkedWith             764
PlatformDesireNextYear         938
FrameworkWorkedWith            216
FrameworkDesireNextYear        363
IDE                            839
OperatingSystem                  4
Methodology                    155
VersionControl                  49
CheckInCode                      6
AIDangerous                      4
AIInteresting                    4
AIResponsible                    4
AIFuture                         3
Age                              6
term                             2
loan_status                      6
pymnt_plan                       1
orignal_website_directory     2946
purpose                         12
title                           12
zip_code                       629
addr_state                      49
dti                           1898
earliest_cr_line               426
last_pymnt_d                    39
last_pymnt_amnt               2678
next_pymnt_d                     2
last_credit_pull_d              40
Time                           698
emp_length                      11
your_favoritearticle_today    2945
url                           2946
Email                         1921
homeaddress                   2629
latitude                      2629
longitude                     2629
PHONE                           24
officeaddress                 2331
website                       1369
dateAdded                       37
previousaddress                 48
phones                          46
CrimeTime                      754
dtype: int64