如何将变量写入 csv 文件而不用单个字符分隔？

Question

My code attempts to split several data tables into year long chunks, then correlate them against all other years and return the correlation values to a matrix. I am attempting to write these outputs to a csv file, and while it is working fine for the matrices themselves, when I try to write the name of the column and table, they are split by their individual characters.

def split_into_yearly_chunks(egauge, column, offset):
      split_into_chunks_stmnt = " SELECT " + column + \
                                " FROM " + egauge + \
                                " OFFSET " + offset + " LIMIT 525600 "
      year_long_chunk = pd.read_sql_query(split_into_chunks_stmnt, engine)
      return year_long_chunk

for x in prof_list:
      for y in col_list:
            list_of_year_long_chunks = []
            for z in off_list:
                  year_long_chunk = split_into_yearly_chunks(x,y,z)
                  if len(year_long_chunk) == 525600:
                        list_of_year_long_chunks.append(year_long_chunk)
            corr_matrix = []
            for corr_year in list_of_year_long_chunks:
                  corr_row = []
                  for corr_partner in list_of_year_long_chunks:
                        corr_value, p_coef = spearmanr(corr_year, corr_partner)
                        corr_row.append(corr_value)
                  corr_matrix.append(corr_row)
            print(x,'; ',y,'; ')
            with open('correlation_data_58_profiles.csv', 'a') as f:
                thewriter = csv.writer(f)
                thewriter.writerow(x)
                thewriter.writerow(y)
            for row in corr_matrix:
                print(row)
                with open('correlation_data_58_profiles.csv', 'a', newline = '') as f:
                    thewriter = csv.writer(f)
                    thewriter.writerow(row)

(Really only the last 10 or so lines in my code are the problem, but I figured I'd give the whole thing for background). My prof,col,and off_lists are all lists of strings. The way that this stores in my csv file looks like this:

e,g,a,u,g,e,1,3,8,3,0

g,r,i,d

1.0,0.7811790818745755,0.7678768782119194,0.7217461539833535
0.7811790818745755,0.9999999999999998,0.7614854144434556,0.714875063672517
0.7678768782119194,0.7614854144434556,0.9999999999999998,0.7215907332829061
0.7217461539833535,0.7148750636725169,0.7215907332829061,0.9999999999999998

I'd like egauge13830 and grid not to be separated by commas, and other answers I've seen wouldn't work for the for loop that I have. How can I do this?

Answer 1

csv.writer(...).writerow expects a list of values, representing the values of a single row.

At some places in the code you are giving it single strings:

thewriter.writerow(x);  # x is a string out of prof_list
thewriter.writerow(y);  # y is a string out of col_list

Since it expects lists of strings, it treats each of these strings as a list of individual characters; that's why you get each character as its own "column", separated by commas.

If you want each of these single strings to appear in its own row as a single column value, then you'll need to make each of them into a one-element-list , indicating to the CSV writer that you want a row consisting of a single value:

thewriter.writerow([x]);  # `[x]` means "a list comprised only of x"
thewriter.writerow([y]);  

---

Also bear in mind that a CSV containing two rows of a single value each, followed by N rows of K values each would be kind of hard to further process; so you should consider if that's really what you want.