illy

AbuAisheh_PhD_2011.html:6:<meta name="Creator" content="PScript5.dll Version 5.2.2">
Abuelela_10_PhD.html:6:<meta name="Creator" content="TeX">
Abuissa09PhD.html:5:<meta name="Creator" content="Acrobat PDFMaker 6.0 for Word">
Acceleration_of_the_DEM_on_a_Reconfigurable_Co-Processor_2003.html:4:<meta name="Creator" content="Acrobat PDFMaker 5.0 for Word">
Adams09PhD.html:6:<meta name="Creator" content="PScript5.dll Version 5.2.2">
Adkins_10_PhD.html:6:<meta name="Creator" content="MicrosoftÆ Office Word 2007">
Ahall_11_PhD.html:6:<meta name="Creator" content="MicrosoftÆ Office Word 2007">
Ahasan07PhD.html:5:<meta name="Creator" content="Acrobat PDFMaker 7.0.7 for Word">
Ahmad10PhD.html:6:<meta name="Creator" content="PScript5.dll Version 5.2">
Ahmad11PhD_A1a.html:6:<meta name="Creator" content="MicrosoftÆ Office Word 2007">

illy

前陣子一直打算謝謝markdown寫作，拖到現在被這位老兄搶了先，很不錯的介紹文章：[为什么Markdown+R有较大概率成为科技写作主流？]（http://www.douban.com/note/220903450/）.所以我就不再重複了，此文打算用問答的形式介紹下markdown對於文科寫作的意義[^ft]。

###1. 什麼是markdown？

markdown是一種輕量化的標記語言。大家所熟知的標記語言還有latex，lyx， XML，HTML，這類文本的特點就是你在編輯時可以使用各種tag來控制文本的格式。話說回來，mac下的pages，M$的docx也都時XML文檔，只是你在編輯的時候，不需要加入tag，軟件本身替你增加tag了。

這類語言最大的好處就是在編輯時，你可以很大程度的減少格式對於文本的干擾。你想啊，打開一個word文檔，你調個文字大小，設置個 加粗，二號標題，再來個 斜體 什麼的, 時間大半過去了。而標記語言的最大優勢在於，你只要使用系統能夠識別的tag就可以非常方便地控制這些格式。比如markdown，你可以先使用系統默認的那些tag來寫作，等到最後輸出的時候，配置下css模版即可。

markdown是各種標記語言中的小兄弟，一來是因為它年輕，二來因為它非常輕量化。md是 John Gruber 和 Aaron Swartz 最初發明的，廣泛用於技術文檔的寫作。它的最大優勢在於人機皆可讀 （machine-readble, human-readble)。

illy

##AWK notes##

1. selective printing

    awk '$2 ~ regex, { $1="", pring $0}' 
   

If $2 = regex, then print the whole line but not $1

2. convert a single line to multiple lines

illy

m.geQuote <- as.matrix(geQuote[,2:5])
acf.geQuote <- acf(m.geQuote, lag=5, plot=F, na.action=na.contiguous)
m.acf.geQuote <- melt(acf.geQuote$acf)

m.acf.geQuote$Var1[m.acf.geQuote$Var1 == 1] <- "Day0"
m.acf.geQuote$Var1[m.acf.geQuote$Var1 == 2] <- "Day1"
m.acf.geQuote$Var1[m.acf.geQuote$Var1 == 3] <- "Day2"
m.acf.geQuote$Var1[m.acf.geQuote$Var1 == 4] <- "Day3"
m.acf.geQuote$Var1[m.acf.geQuote$Var1 == 5] <- "Day4"

illy

m.acf.geQuote$Var1[m.acf.geQuote$Var1 == 1] <- "Day0"
m.acf.geQuote$Var1[m.acf.geQuote$Var1 == 2] <- "Day1"
m.acf.geQuote$Var1[m.acf.geQuote$Var1 == 3] <- "Day2"
m.acf.geQuote$Var1[m.acf.geQuote$Var1 == 4] <- "Day3"
m.acf.geQuote$Var1[m.acf.geQuote$Var1 == 5] <- "Day4"
m.acf.geQuote$Var1 <- factor(m.acf.geQuote$Var1, 
+ levels=unique(m.acf.geQuote$Var1), ordered=T)

m.acf.geQuote$Var2[m.acf.geQuote$Var2 == 1] <- "Open"
m.acf.geQuote$Var2[m.acf.geQuote$Var2 == 2] <- "Close"

illy

p <- ggplot(m.acf.geQuote)
p <- p + geom_raster(aes(x=Var1, y=Var2, lable=value, fill= value)) +
  facet_wrap(~Var3, nrow=4) +
  ggtitle("Cross-correlation of 4 different prices of GE ticker") +
  theme(legend.position="none") +
  labs(fill="Correlation") +
  xlab("") + ylab("")
print(p) 

illy

str(acf.geQuote)

List of 6
 $ acf   : num [1:5, 1:4, 1:4] 1 -0.1917 -0.478 0.1049 0.0648 ...
 $ type  : chr "correlation"
 $ n.used: int 5
 $ lag   : num [1:5, 1:4, 1:4] 0 1 2 3 4 0 -1 -2 -3 -4 ...
 $ series: chr "m.geQuote"
 $ snames: chr [1:4] "Open" "Close" "Low" "High"
 - attr(*, "class")= chr "acf"

illy

m.geQuote <- as.matrix(geQuote[,2:5])
acf.geQuote <- acf(m.geQuote, lag=5, plot=F, na.action=na.contiguous)
m.acf.geQuote <- melt(acf.geQuote$acf)

illy

## This script is for crawling tweets with a specific address file.


#!/usr/bin/env bash

DIR=PARENT_DIR/`date "+%d-%m-%y-%H:%M"` #set the download file based to download date
mkdir -p $DIR #make dir according to above

wget -i EXTERNAL_ADDRESS_LIST -np -r -N -l1 -P $DIR 

illy

##1 Data manipulation

1. If the data contains NA values, it regards it as factor, not numeric.

    DATA$COLUMN <- as.numeric(as.character(DATA$COLUMN))
   

2. Rename the column:

    names(DATA)[2] <- "NEW_NAME"