extracting rows from CSV file based on specific keywords

Question

enter image description hereI have created a code to help me retrieving the data from csv file

  import re
keywords = {"metal", "energy", "team", "sheet", "solar" "financial", "transportation", "electrical", "scientists",
            "electronic", "workers"}  # all your keywords


keyre=re.compile("energy",re.IGNORECASE)
with open("2006-data-8-8-2016.csv") as infile:
    with open("new_data.csv", "w") as outfile:
        outfile.write(infile.readline())  # Save the header
        for line in infile:
            if len(keyre.findall(line))>0:
                outfile.write(line)

I need it to look for each keyword in two main columns which are "position" and "Job description" , and then take the whole row that includes these words and write them in the new file. Any ideas on how this can be done in the simplest way?

Answer 1

Try this, looping in a dataframe and write back a new dataframe to a csv file.

import pandas as pd

keywords = {"metal", "energy", "team", "sheet", "solar", "financial", 
        "transportation", "electrical", "scientists",
        "electronic", "workers"}  # all your keywords

df = pd.read_csv("2006-data-8-8-2016.csv", sep=",")

listMatchPosition = []
listMatchDescription = []

for i in range(len(df.index)):
    if any(x in df['position'][i] or x in df['Job description'][i] for x in keywords):
        listMatchPosition.append(df['position'][i])
        listMatchDescription.append(df['Job description'][i])


output = pd.DataFrame({'position':listMatchPosition, 'Job description':listMatchDescription})
output.to_csv("new_data.csv", index=False)

EDIT: If you have many columns to add, the modified following code will do the job.

df = pd.read_csv("2006-data-8-8-2016.csv", sep=",")

output = pd.DataFrame(columns=df.columns)

for i in range(len(df.index)):
    if any(x in df['position'][i] or x in df['Job description'][i] for x in keywords):
    output.loc[len(output)] = [df[j][i] for j in df.columns]

output.to_csv("new_data.csv", index=False)

Answer 2

You can do this using pandas as follows, if you are looking for rows that contain exactly one word from the list of keywords:

keywords = ["metal", "energy", "team", "sheet", "solar" "financial", "transportation", "electrical", "scientists",
            "electronic", "workers"]

# read the csv data into a dataframe 
# change "," to the data separator in your csv file 
df = pd.read_csv("2006-data-8-8-2016.csv", sep=",")
# filter the data: keep only the rows that contain one of the keywords 
# in the position or the Job description columns
df = df[df["position"].isin(keywords) | df["Job description"].isin(keywords)] 
# write the data back to a csv file 
df.to_csv("new_data.csv",sep=",", index=False) 

If you are looking for substrings in the rows (e.g looking financial in financial engineering) then you can do the following:

keywords = ["metal", "energy", "team", "sheet", "solar" "financial", "transportation", "electrical", "scientists",
            "electronic", "workers"]
searched_keywords = '|'.join(keywords)

# read the csv data into a dataframe 
# change "," to the data separator in your csv file 
df = pd.read_csv("2006-data-8-8-2016.csv", sep=",")
# filter the data: keep only the rows that contain one of the keywords 
# in the position or the Job description columns
df = df[df["position"].str.contains(searched_keywords) | df["Job description"].str.contains(searched_keywords)] 
# write the data back to a csv file 
df.to_csv("new_data.csv",sep=",", index=False)