R中的.csv数据操作而不是python

Question

I have a simple .csv format data that need to be manipulated first before I able to create the plot accords to this data. However, I understand how to manipulate .csv format data from python. I want to apply same logic in R but I am not sure how to do this.

Below is the example data from .csv file but load into R. I have created the code for us to discuss this issue.

df <- data.frame(Name = c("AC", "AC", "PT", "PT", "OR", "OR"),
    useless_column = c("","","A",3,4," "),
  measurement = c("H", "", "K", "M", "", "H"),
  amount = c(12, 54, 20, 87, 75, 22),
    useless_column = c("","","A",3,4," ")) 

In python, I generally will do this:

import csv
import os
import glob
import sys
fileList = glob.glob("R:xxxxxxxxxxxxxxxxxxxxx\*.csv")
for inputFile in fileList:
        outputFilename = inputFile + "output.csv"
        csvInput = csv.reader(open(inputFile,'r'),delimiter=",")
        outputFile = open(outputFilename,'w')
        outputFile.write("Name,measurement,amount\n")
        csvInput.next()
        for line in csvInput:
            if line[2] == "H":
               meas = "100"
            elif line[2] == "K":
               meas = "1000"
            elif line[2] == "M":
               meas = "1000000"
            else:
               meas = "1"
            amount = int(meas) * line[3]

            outputFile.write(",".join(line[0],line[2],amount+"\n"]))
outputFile.close()

In python, I can load the csv and then using for loop to identify of each line from the csv file. Then tailor-made my output file before I continue my analysis. From above, I expect my output something like below and the code is in R format:

    df <- data.frame(Name = c("AC", "AC", "PT", "PT", "OR", "OR"),
  measurment = c("H", "", "K", "M", "", "H"),
  amount = c(1200, 54, 20000, 87000000, 75, 2200))

I would like to know to do this in R? I have a small code of R and plese anyone can guide me into the correct direction:

x <- read.csv("xxxx.csv", header=T,sep=",")
xC = ncol(x)
xR = nrow(x)
op = data.frame(matrix(data = x, nrow= xR, ncol=3,byrow=T))
for (x in :xC)
{
    for (r in 1:xR)
    {
    xxxxxxxx

    }

Answer 1

Adapting python code in R means giving up the loops in favor of vectorized operations. Here, we can create meas based on a named vector, and then compute amount:

# dictionnary of measurement values:
m <- c(H = 100, K = 1000, M = 1000000)

# create meas based on measurement
df$meas <- m[df$measurment]
df$meas[is.na(df$meas)] <- 1
# compute amount
df$amount <- df$meas * df$amount

Data

df <- data.frame(Name = c("AC", "AC", "PT", "PT", "OR", "OR"),
                 measurment = c("H", "", "K", "M", "", "H"),
                 amount = c(1200, 54, 20000, 87000000, 75, 2200))

Answer 2

Have you tried using pandas.read_csv? Or are the csv files so irregular that you cannot use pandas' read_csv method to read them?

You can do a for loop to manipulate your data from each file, and then append it to a master DataFrame .

Example:

import pandas as pd

PATH = '/home/data/' # Example path

master_df = pd.DataFrame()
for inputFile in fileList:
    csv_file = pd.read_csv(path + inputFile, sep=',')
    H_index = csv_file[csv_file.loc[:, 2] == 'H'].index
    csv_file.loc[H_index, 3] = csv_file.loc[H_index, 3] * 100
    master_df = master_df.append(csv_file)

I've skipped the K and M part of the manipulation.

You could directly plot from master_df by doing something like

master_df.plot()

Answer 3

You've got the code to read in the data (read.csv), so am I right in thinking your main struggle is in the manimpuation itself?

If so, you could carry on using lots if and for loops, but I think there are much easier ways. Something like:

df <- read.csv("xxxx.csv", header=T,sep=",")
df$meas <- df$measurement # Create a new column called 'meas' by copying column 'measurement'
df$meas[df$meas == "H"] <- 100 # Replace H's with 100
df$meas[df$meas == "K"] <- 1000
df$meas[df$meas == "M"] <- 1000000
df$value <- df$meas * df$amount