python blaze (pandas) 无法 safley 转换 <i8 的用户 dtype

Question

我想使用 python blaze 从uk nga geonames download 读取uk.txt 文件，然后使用odo 将其插入到 Postgresql 数据库中。

代码是：

import blaze as bz
from odo import odo

dataPath = 'uk.txt'
myData = bz.Data(dataPath, sep='\t')
out = odo(myData, 'postgresql://postgres:postgres@localhost:5432/blaze_test::uk_geonames')

我收到错误ValueError: cannot safely convert passed user dtype of <i8 for object dtyped data in column 0，我认为我理解为“无法转换数据类型以插入数据库”

我应该强制dtype 等于什么吗？我该如何解决这个问题？

来自文件的示例输入是：

RC  UFI UNI LAT LONG    DMS_LAT DMS_LONG    MGRS    JOG FC  DSG PC  CC1 ADM1    POP ELEV    CC2 NT  LC  SHORT_FORM  GENERIC SORT_NAME_RO    FULL_NAME_RO    FULL_NAME_ND_RO SORT_NAME_RG    FULL_NAME_RG    FULL_NAME_ND_RG NOTE    MODIFY_DATE DISPLAY NAME_RANK   NAME_LINK   TRANSL_CD   NM_MODIFY_DATE

1   380952  475802  54.086111   -6.655556   540510  -63920  29UPV5334795644 NN29-06 H   STM     EI,UK               EI,UK   N       Clarebane       CLAREBANERIVER  Clarebane River Clarebane River CLAREBANERIVER  Clarebane River Clarebane River     2014-06-27  1,2,3   2

Answer 1

由于某种原因，无法正确推断标题。您可以像这样传入infer_header 关键字参数：

In [12]: from blaze import Data

In [13]: from odo import CSV, odo

In [14]: d = Data(CSV('uk.txt', sep='\t', has_header=True))

In [15]: d.head(5)
Out[15]:
   RC     UFI     UNI        LAT      LONG  DMS_LAT  DMS_LONG  \
0   1  380952  475802  54.086111 -6.655556   540510    -63920
1   1  380952  475801  54.086111 -6.655556   540510    -63920
2   1  380954  475805  54.104722 -6.648889   540617    -63856
3   1  380955  475806  54.098056 -6.644167   540553    -63839
4   1  380958  475810  54.040556 -6.614444   540226    -63652

              MGRS      JOG FC      ...          SORT_NAME_RG  \
0  29UPV5334795644  NN29-06  H      ...        CLAREBANERIVER
1  29UPV5334795644  NN29-06  H      ...             CLAREBANE
2  29UPV5371497729  NN29-06  H      ...           ALINA LOUGH
3  29UPV5404796997  NN29-06  H      ...          CORLISSLOUGH
4  29UPV5620690667  NN29-06  H      ...          DRUMBOYLOUGH

      FULL_NAME_RG  FULL_NAME_ND_RG NOTE MODIFY_DATE DISPLAY NAME_RANK  \
0  Clarebane River  Clarebane River  NaN  2014-06-27   1,2,3         2
1        Clarebane        Clarebane  NaN  2014-06-27   1,2,3         1
2     Alina, Lough     Alina, Lough  NaN  2014-06-27   1,2,3         1
3    Corliss Lough    Corliss Lough  NaN  2014-06-27   1,2,3         1
4    Drumboy Lough    Drumboy Lough  NaN  2014-06-27   1,2,3         1

  NAME_LINK TRANSL_CD NM_MODIFY_DATE
0       NaN       NaN     2014-06-27
1       NaN       NaN     2014-06-27
2       NaN       NaN     2014-06-27
3       NaN       NaN     2014-06-27
4       NaN       NaN     2014-06-27

[5 rows x 34 columns]

之后，只需将odo 放入所需的表中：

In [16]: t = odo(d, 'postgresql://localhost::uk')

In [17]: uk = Data(t)

In [19]: uk.head(5)
Out[19]:
   RC     UFI     UNI        LAT      LONG  DMS_LAT  DMS_LONG  \
0   1  380952  475802  54.086111 -6.655556   540510    -63920
1   1  380952  475801  54.086111 -6.655556   540510    -63920
2   1  380954  475805  54.104722 -6.648889   540617    -63856
3   1  380955  475806  54.098056 -6.644167   540553    -63839
4   1  380958  475810  54.040556 -6.614444   540226    -63652

              MGRS      JOG FC      ...          SORT_NAME_RG  \
0  29UPV5334795644  NN29-06  H      ...        CLAREBANERIVER
1  29UPV5334795644  NN29-06  H      ...             CLAREBANE
2  29UPV5371497729  NN29-06  H      ...           ALINA LOUGH
3  29UPV5404796997  NN29-06  H      ...          CORLISSLOUGH
4  29UPV5620690667  NN29-06  H      ...          DRUMBOYLOUGH

      FULL_NAME_RG  FULL_NAME_ND_RG NOTE MODIFY_DATE DISPLAY NAME_RANK  \
0  Clarebane River  Clarebane River  NaN  2014-06-27   1,2,3         2
1        Clarebane        Clarebane  NaN  2014-06-27   1,2,3         1
2     Alina, Lough     Alina, Lough  NaN  2014-06-27   1,2,3         1
3    Corliss Lough    Corliss Lough  NaN  2014-06-27   1,2,3         1
4    Drumboy Lough    Drumboy Lough  NaN  2014-06-27   1,2,3         1

  NAME_LINK TRANSL_CD NM_MODIFY_DATE
0       NaN       NaN     2014-06-27
1       NaN       NaN     2014-06-27
2       NaN       NaN     2014-06-27
3       NaN       NaN     2014-06-27
4       NaN       NaN     2014-06-27

[5 rows x 34 columns]