python - Pandas 可以将可变长度空格作为列分隔符处理吗

Question

这个问题在这里已经有了答案:

How to make separator in pandas read_csv more flexible wrt whitespace, for irregular separators? (4 个回答)

关闭 5年前。

我有一个文本文件，其中的列由可变数量的空格分隔。是否可以在不预处理文件的情况下直接将此文件作为 Pandas 数据框加载？在 pandas documentation the delimiter section说我可以使用 's*' 构造，但我无法让它工作。

## sample data
head sample.txt

#                                                                            --- full sequence --- -------------- this domain -------------   hmm coord   ali coord   env coord
# target name        accession   tlen query name           accession   qlen   E-value  score  bias   #  of  c-Evalue  i-Evalue  score  bias  from    to  from    to  from    to  acc description of target
#------------------- ---------- ----- -------------------- ---------- ----- --------- ------ ----- --- --- --------- --------- ------ ----- ----- ----- ----- ----- ----- ----- ---- ---------------------
ABC_membrane         PF00664.18   275 AAF67494.2_AF170880  -            615     8e-29  100.7  11.4   1   1     3e-32     1e-28  100.4   7.9     3   273    42   313    40   315 0.95 ABC transporter transmembrane region
ABC_tran             PF00005.22   118 AAF67494.2_AF170880  -            615   2.6e-20   72.8   0.0   1   1   1.9e-23   6.4e-20   71.5   0.0     1   118   402   527   402   527 0.93 ABC transporter
SMC_N                PF02463.14   220 AAF67494.2_AF170880  -            615   3.8e-08   32.7   0.2   1   2    0.0036        12    4.9   0.0    27    40   391   404   383   408 0.86 RecF/RecN/SMC N terminal domain
SMC_N                PF02463.14   220 AAF67494.2_AF170880  -            615   3.8e-08   32.7   0.2   2   2   1.8e-09   6.1e-06   25.4   0.0   116   210   461   568   428   575 0.85 RecF/RecN/SMC N terminal domain
AAA_16               PF13191.1    166 AAF67494.2_AF170880  -            615   3.1e-06   27.5   0.3   1   1     2e-09     7e-06   26.4   0.2    20   158   386   544   376   556 0.72 AAA ATPase domain
YceG                 PF02618.11   297 AAF67495.1_AF170880  -            284   3.4e-64  216.6   0.0   1   1   2.9e-68     4e-64  216.3   0.0    68   296    53   274    29   275 0.85 YceG-like family
Pyr_redox_3          PF13738.1    203 AAF67496.2_AF170880  -            352   2.9e-28   99.1   0.0   1   2   2.8e-30   4.8e-27   95.2   0.0     1   201     4   198     4   200 0.85 Pyridine nucleotide-disulphide oxidoreductase

#load data
from pandas import *
data = read_table('sample.txt', skiprows=3, header=None, sep=" ")

ValueError: Expecting 83 columns, got 91 in row 4

#load data part 2
data = read_table('sample.txt', skiprows=3, header=None, sep="'s*' ")
#this mushes some of the columns into the first column and drops the rest.
    X.1
1    ABC_tran PF00005.22 118 AAF67494.2_
2    SMC_N PF02463.14 220 AAF67494.2_
3    SMC_N PF02463.14 220 AAF67494.2_
4    AAA_16 PF13191.1 166 AAF67494.2_
5    YceG PF02618.11 297 AAF67495.1_
6    Pyr_redox_3 PF13738.1 203 AAF67496.2_
7    Pyr_redox_3 PF13738.1 203 AAF67496.2_
8    FMO-like PF00743.14 532 AAF67496.2_
9    FMO-like PF00743.14 532 AAF67496.2_

虽然我可以预处理文件以将空格更改为逗号/制表符，但直接加载它们会很好。

(仅供引用，这是 hmmscan program 的 *.hmmdomtblout 输出)

Answer 1

你应该能够做到这一点，@DSM 刚刚在另一个帖子中教过我:

data = read_table('sample.txt', skiprows=3, header=None, delim_whitespace=True)

Documentation