Created
December 19, 2015 06:40
-
-
Save rldotai/ec1a955b870390718591 to your computer and use it in GitHub Desktop.
A quick script to format tables that were in the `verbatim` environment as proper tables using Pandas.
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
#!python3 | |
""" | |
A script to read a document containing tables that I left as "verbatim" and | |
format them properly in LaTeX. | |
We go from this: | |
\begin{verbatim} | |
sparsity num_active accuracy test_accuracy train_accuracy | |
count 4000.000000 4000.000000 4000.000000 4000.000000 4000 | |
mean 0.984240 124.657750 0.957216 0.829739 1 | |
std 0.000943 7.460431 0.020602 0.079698 0 | |
min 0.981290 94.000000 0.864583 0.500000 1 | |
25% 0.983565 119.000000 0.947917 0.772727 1 | |
50% 0.984197 125.000000 0.958333 0.833333 1 | |
75% 0.984956 130.000000 0.968750 0.875000 1 | |
max 0.988116 148.000000 1.000000 1.000000 1 | |
\end{verbatim} | |
to this: | |
\begin{tabular}{lrrrrr} | |
\toprule | |
{} & sparsity & num\_active & accuracy & test\_accuracy & train\_accuracy \\ | |
\midrule | |
mean & 0.9842 & 124.6577 & 0.9572 & 0.8297 & 1.0000 \\ | |
std & 0.0009 & 7.4604 & 0.0206 & 0.0797 & 0.0000 \\ | |
min & 0.9813 & 94.0000 & 0.8646 & 0.5000 & 1.0000 \\ | |
25\% & 0.9836 & 119.0000 & 0.9479 & 0.7727 & 1.0000 \\ | |
50\% & 0.9842 & 125.0000 & 0.9583 & 0.8333 & 1.0000 \\ | |
75\% & 0.9850 & 130.0000 & 0.9688 & 0.8750 & 1.0000 \\ | |
max & 0.9881 & 148.0000 & 1.0000 & 1.0000 & 1.0000 \\ | |
\bottomrule | |
\end{tabular} | |
""" | |
import pandas as pd | |
import sys | |
def read_table(text, separator=None, linebreak='\n', header=True, index=True): | |
def parse_row(line): | |
return [x.strip() for x in line.split(separator)] | |
rows = text.split(linebreak) | |
# name the columns according to the first row, if desired | |
if header: | |
cols = parse_row(rows.pop(0)) | |
else: | |
cols = None | |
# parse each row | |
data = [parse_row(x) for x in rows] | |
# if we want to keep the index separate, we can do so | |
if index: | |
indices = [x.pop(0) for x in data] | |
else: | |
indices = None | |
# convert to a pandas dataframe | |
return pd.DataFrame(data, columns=cols, index=indices) | |
def my_df_format(df): | |
"""Quick and dirty formatting of a dataframe. Adapt as necessary.""" | |
# sparsity num_active accuracy test_accuracy, train_accuracy | |
df = df.copy() | |
df = df.drop(['count']) | |
dtypes = { | |
'sparsity': float, | |
'num_active': float, | |
'accuracy': float, | |
'test_accuracy': float, | |
'train_accuracy': float, | |
} | |
for k, v in dtypes.items(): | |
df[k] = df[k].astype(v) | |
formats = { | |
'sparsity': '{:.4f}'.format, | |
'num_active': '{:.4f}'.format, | |
'accuracy': '{:.4f}'.format, | |
'test_accuracy': '{:.4f}'.format, | |
'train_accuracy': '{:.4f}'.format, | |
} | |
# print(df.to_latex(formatters=formats)) | |
return df.to_latex(formatters=formats) | |
def format_document(filename): | |
"""Process the document; a bit of a kludge.""" | |
with open(filename, 'r') as f: | |
lines = f.readlines() | |
ret = [] | |
in_table = False | |
for line in lines: | |
# print(line) | |
if r'\begin{verbatim}' in line: | |
if in_table: | |
raise Exception('Looks like something went wrong') # exceptional exception | |
else: | |
in_table = True | |
tmp = [] | |
elif in_table: | |
if r'\end{verbatim}' in line: | |
txt = ''.join(tmp).strip() | |
# print(txt) | |
df = read_table(txt) | |
ret.append(my_df_format(df)) | |
in_table = False | |
else: | |
tmp.append(line) | |
else: | |
ret.append(line) | |
return ''.join(ret) | |
# Get the name of the file to operate on, and optionally an output | |
if __name__ == "__main__": | |
txt = format_document(sys.argv[1]) | |
if len(sys.argv) > 2: | |
open(sys.argv[2], 'w').write(txt) | |
else: | |
print(txt) | |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment