Add excel write to analysis_mysql

analysis_mysql.py

@@ -1,6 +1,7 @@
 import argparse
 from typing import Tuple
 import numpy as np
+import xlsxwriter
 from experimentation.Sets import Datasets
 from experimentation.Utils import TextColor
 from experimentation.Database import MySQL
@@ -71,6 +72,14 @@ def parse_arguments() -> Tuple[str, str, str, bool, bool]:
         required=False,
         default=False,
     )
+    ap.add_argument(
+        "-x",
+        "--excel",
+        type=str,
+        default="",
+        required=False,
+        help="generate excel file",
+    )
     args = ap.parse_args()
     return (
         args.experiment,
@@ -79,9 +88,45 @@ def parse_arguments() -> Tuple[str, str, str, bool, bool]:
         args.tex_output,
         args.compare,
         args.time,
+        args.excel,
     )
 
 
+def excel_write_header(book, sheet):
+    header = book.add_format()
+    header.set_font_size(18)
+    subheader = book.add_format()
+    subheader.set_font_size(16)
+    bold = book.add_format({"bold": True, "font_size": 14})
+    sheet.write(0, 0, "Dataset", bold)
+    sheet.set_column(0, 0, 30)
+    i = 1
+    lengths = [10, 10, 10, 10, 10, 10, 10]
+    for item, length in zip(models_tree, lengths):
+        sheet.write(0, i, item, bold)
+        sheet.set_column(i, i, length)
+        i += 1
+
+
+def excel_write_line(book, sheet, dataset, line):
+    try:
+        excel_write_line.row += 1
+    except AttributeError:
+        excel_write_line.row = 1
+    size_n = 14
+    decimal = book.add_format({"num_format": "0.000000", "font_size": size_n})
+    normal = book.add_format({"font_size": size_n})
+    col = 0
+    excel_ws.write(excel_write_line.row, col, dataset, normal)
+    for item in line.values():
+        sheet.write(excel_write_line.row, col + 1, item, decimal)
+        col += 1
+
+
+def excel_write_footer(book):
+    book.close()
+
+
 def print_header_tex(file_tex, second=False):
     # old_header = (
     #     "\\begin{table}[ht]\n"
@@ -209,6 +254,7 @@ def report_footer(agg):
     tex_output,
     compare,
     time_info,
+    excel,
 ) = parse_arguments()
 dbh = MySQL()
 database = dbh.get_connection()
@@ -240,6 +286,11 @@ for item in models:
     fields += (f"{item}",)
 report_header(title, experiment, model_type)
 color = TextColor.LINE1
+if excel != "":
+    file_name = f"{excel}.xlsx"
+    excel_wb = xlsxwriter.Workbook(file_name)
+    excel_ws = excel_wb.add_worksheet("exreport")
+    excel_write_header(excel_wb, excel_ws)
 agg = {}
 for item in [
     "better",
@@ -266,6 +317,7 @@ for number, dataset in enumerate(dt):
     line["leaves"] = 0.0
     line["depth"] = 0.0
     line_tex = line.copy()
+    line_excel = {}
     for column, model in enumerate(models):
         record = dbh.find_best(dataset[0], model, experiment)
         if record is None:
@@ -291,6 +343,10 @@ for number, dataset in enumerate(dt):
                 line[model] = color + item
             if csv_output:
                 print(f"{dataset[0]}, {model}, {accuracy}", file=file_csv)
+            if excel != "":
+                line_excel[model] = accuracy
+    if excel != "":
+        excel_write_line(excel_wb, excel_ws, dataset[0], line_excel)
     if tex_output:
         print_line_tex(number + 1, dataset[0], line_tex, file_tex)
         if number == 24:
@@ -303,6 +359,8 @@ for number, dataset in enumerate(dt):
             TextColor.LINE2 if color == TextColor.LINE1 else TextColor.LINE1
         )
         print(report_line(line))
+if excel != "":
+    excel_write_footer(excel_wb)
 report_footer(agg)
 if csv_output:
     file_csv.close()

requirements.txt

@@ -5,3 +5,4 @@ ipympl
 mysql-client
 mysql-connector
 sshtunnel
+xslxwriter