func GenerateData(configurationFile string) {
if configurationFile == "" {
panic("I can't work without a configuration file")
}
log.Info("Loading config")
conf, err := config.LoadConfig(configurationFile)
if err != nil {
panic(err)
}
//api := GetTwitter(&conf)
db := nutz.NewStorage(configurationFile+".db", 0600, nil)
mygraph := Graph{Nodes: []Node{}, Links: []Link{}}
innercount := 0
nodecount := 0
group := 0
for _, account := range conf.TwitterAccounts {
tweets := db.Get(account, "tweets")
from := db.Get(account, "from")
retweets := db.Get(account, "retweets")
unique_mentions := db.Get(account, "unique_mentions")
total_mentions := db.Get(account, "total_mentions")
followers := db.Get(account, "followers")
following := db.Get(account, "following")
followers_followed := db.Get(account, "followers_followed")
mentions_to_followed := db.Get(account, "mentions_to_followed")
log.Info("Account: " + account)
log.Info("from: " + string(from.Data))
log.Info("Tweets: " + string(tweets.Data))
log.Info("retweets: " + string(retweets.Data))
log.Info("unique_mentions: " + string(unique_mentions.Data))
log.Info("total_mentions: " + string(total_mentions.Data))
log.Info("followers: " + string(followers.Data))
log.Info("following: " + string(following.Data))
log.Info("followers_followed: " + string(followers_followed.Data))
log.Info("mentions_to_followed: " + string(mentions_to_followed.Data))
myUniqueMentions := db.GetAll(account, "map_unique_mentions").DataList
nUniqueMentions, _ := strconv.Atoi(string(unique_mentions.Data))
nMentions_to_followed, _ := strconv.Atoi(string(mentions_to_followed.Data))
nTweets, _ := strconv.Atoi(string(tweets.Data))
nReTweets, _ := strconv.Atoi(string(retweets.Data))
om := OutsideMentions(nUniqueMentions, nMentions_to_followed)
apt := AnswerPeopleTax(nUniqueMentions, nMentions_to_followed, nTweets, nReTweets)
if math.IsNaN(float64(om)) {
om = float32(0.01)
}
if math.IsNaN(float64(apt)) {
apt = float32(0.01)
}
// fmt.Println("\tDemocracy tax: " + FloatToString(di))
fmt.Println("\tOutside of circle mentions: " + FloatToString(om))
fmt.Println("\t of answering to external people: " + FloatToString(apt))
mygraph.Nodes = append(mygraph.Nodes, Node{Name: account, Group: group, Thickness: om, Size: apt})
for k, v := range myUniqueMentions {
//id, _ := strconv.ParseInt(k, 10, 64)
//User, _ := api.GetUsersShowById(id, nil)
//log.Info("[" + User.ScreenName + "]:" + string(v))
// now you can put User.ScreeName in the name of the node
weight, _ := strconv.Atoi(string(v))
mygraph.Nodes = append(mygraph.Nodes, Node{Name: string(k), Group: group, Thickness: 0.01, Size: 0.01})
mygraph.Links = append(mygraph.Links, Link{Source: innercount, Target: nodecount, Value: weight})
innercount++
}
innercount++
nodecount = innercount
group++
}
fileJson, _ := json.MarshalIndent(mygraph, "", " ")
err = ioutil.WriteFile(configurationFile+".output", fileJson, 0644)
if err != nil {
log.Info("WriteFileJson ERROR: " + err.Error())
}
}
func GenerateData(configurationFile string) {
if configurationFile == "" {
panic("I can't work without a configuration file")
}
log.Info("Loading config")
conf, err := config.LoadConfig(configurationFile)
if err != nil {
panic(err)
}
//api := GetTwitter(&conf)
fmt.Println(">> Exporting gathered data")
mygraph := Graph{Nodes: []Node{}, Links: []Link{}, Mode: "static", Defaultedgetype: "undirected"}
for _, account := range conf.TwitterAccounts {
fmt.Println(">> Generating graph for " + account)
mygraph = Graph{Nodes: []Node{}, Links: []Link{}, Mode: "static", Defaultedgetype: "undirected"}
db := nutz.NewStorage(account+".db", 0600, nil)
nodecount := 0
myNetwork := db.GetAll(account, MATCHING_HASHTAGS).DataList
myMatrix := make(map[string][]int) // this is the Matrix Hashtags/ Users ID
myNetworkMatrix := make(map[string]map[string]int8) //so we can extract later data easily
myMapNetwork := make(map[int]string) //this will be used to resolve User ID of the graph <-> Twitter id
var myCSV [][]string
HashtagsMap := db.GetAll(account, GENERATED_HASHTAGS)
var Hashtags []string
Hashtags = append(Hashtags, "UserID") //First column reserved to userid
for h, _ := range HashtagsMap.DataList {
Hashtags = append(Hashtags, string(h))
}
myCSV = append(myCSV, Hashtags)
for k, _ := range myNetwork {
// ki64, _ := strconv.ParseInt(string(k), 10, 64)
ki64 := string(k)
//Column name is ki64
myUserNetwork := db.GetAll(account, MATCHING_HASHTAGS, k).DataList
var userOccurrence []string
userOccurrence = append(userOccurrence, string(k)) //this is the userid
for _, h := range Hashtags {
if occurence, ok := myUserNetwork[h]; ok {
userOccurrence = append(userOccurrence, string(occurence))
} else {
userOccurrence = append(userOccurrence, strconv.Itoa(0))
}
}
myCSV = append(myCSV, userOccurrence)
myNetworkMatrix[ki64] = make(map[string]int8)
//fmt.Println("User ID is: " + string(k))
if len(myUserNetwork) > conf.HashtagCutOff { //Cutting off people that just tweeted 1 hashtag
htagsatisfied := false //cutoff var
for h, o := range myUserNetwork {
occur, _ := strconv.Atoi(string(o))
//fmt.Println("Hastag: " + string(h) + " saw " + string(o) + " times")
if occur > conf.HashtagOccurrenceCutOff { // Cutting off people that just tweeted the hashtag once
myMatrix[string(h)] = append(myMatrix[string(h)], nodecount) // Generating adjacient map
htagsatisfied = true //cutoff var, setting true to enable node counting
}
occurrences, _ := strconv.Atoi(string(o))
myNetworkMatrix[ki64][string(h)] = int8(occurrences) // convert the db to a map
}
if htagsatisfied { //Cutting off also nodes who satisfied the cuttoff above
myMapNetwork[nodecount] = ki64 //mapping Graph user id with Tweet user id
mygraph.Nodes = append(mygraph.Nodes, Node{Id: nodecount, Name: string(k), Group: 1})
nodecount++
}
}
}
fmt.Println(">> Preparing graph for " + account)
linkscount := 0
for hashtag, users := range myMatrix {
for _, userid := range users {
for _, userid2 := range users {
if userid2 != userid {
if int(myNetworkMatrix[myMapNetwork[userid]][hashtag]) > conf.HashtagOccurrenceCutOff {
mygraph.Links = append(mygraph.Links, Link{Id: linkscount, Source: userid, Target: userid2, Value: float32(myNetworkMatrix[myMapNetwork[userid]][hashtag])})
linkscount++
}
}
}
}
}
fmt.Println(">> Writing matrix to csv")
utils.WriteCSV(myCSV, account+".csv")
fmt.Println(">> Writing graph to json file")
//
// nUniqueMentions, _ := strconv.Atoi(string(unique_mentions.Data))
// nMentions_to_followed, _ := strconv.Atoi(string(mentions_to_followed.Data))
// nTweets, _ := strconv.Atoi(string(tweets.Data))
// nReTweets, _ := strconv.Atoi(string(retweets.Data))
//mygraph.Nodes = append(mygraph.Nodes, Node{Name: account, Group: group})
// for k, v := range myUniqueMentions {
// weight, _ := strconv.Atoi(string(v))
// mygraph.Nodes = append(mygraph.Nodes, Node{Name: string(k), Group: group, Thickness: 0.01, Size: 0.01})
// mygraph.Links = append(mygraph.Links, Link{Source: innercount, Target: nodecount, Value: weight})
// innercount++
// }
fileJson, _ := ffjson.Marshal(&mygraph)
err = ioutil.WriteFile(account+".output", fileJson, 0644)
if err != nil {
log.Info("WriteFileJson ERROR: " + err.Error())
}
out, err := xml.Marshal(mygraph)
out = append([]byte(`<?xml version="1.0" encoding="UTF-8"?><gexf xmlns="http://www.gexf.net/1.2draft" version="1.2"> <meta lastmodifieddate="2009-03-20"> <creator>dark-lab</creator><description>Gephi file</description> </meta>`), out...)
out = append(out, []byte(`</gexf>`)...)
err = ioutil.WriteFile(account+".output.gexf", out, 0644)
if err != nil {
log.Info("WriteFileJson ERROR: " + err.Error())
}
}
}
func GatherData(configurationFile string) {
if configurationFile == "" {
panic("I can't work without a configuration file")
}
log.Info("Loading config")
conf, err := config.LoadConfig(configurationFile)
if err != nil {
panic(err)
}
myTweets := make(map[string]timelinesTweets)
api := GetTwitter(&conf)
db := nutz.NewStorage(configurationFile+".db", 0600, nil)
retweetRegex, _ := regexp.Compile(`^RT`) // detecting retweets
for _, account := range conf.TwitterAccounts {
log.Info("-== Timeline for Account: %#v ==-\n", account)
myTweets[account] = GetTimelines(api, account, conf.FetchFrom)
log.Info("-== END TIMELINE for %#v ==-\n", account)
}
log.Info("Analyzing && collecting data")
for i := range myTweets {
var retweets int
var mymentions int
var mentions struct {
Name string
Indices []int
Screen_name string
Id int64
Id_str string
}
var myUniqueMentions map[int64]int
myUniqueMentions = make(map[int64]int)
fmt.Println("-== Account: " + i + " ==-")
fmt.Println("\tTweets: " + strconv.Itoa(len(myTweets[i])))
for _, t := range myTweets[i] {
// detecting retweets
if retweetRegex.MatchString(t.Text) == true {
retweets++
} else {
//detecting mentions outside retweets
for _, mentions = range t.Entities.User_mentions {
mymentions++
if t.InReplyToUserID != 0 { //we are interested only in replies
myUniqueMentions[mentions.Id]++
}
}
}
}
if conf.FetchFollow == true {
log.Info("-== GetFollowers for Account: %#v ==-\n", i)
Followers := GetFollowers(api, i)
log.Info("-== GetFollowing for Account: %#v ==-\n", i)
Following := GetFollowing(api, i)
log.Info("-== End getting Following/Followers for Account: %#v ==-\n", i)
var Corrispective []int64
var MentionsWithCorrispective []int64
for _, i := range Following {
if _, ok := myUniqueMentions[i]; ok {
MentionsWithCorrispective = append(MentionsWithCorrispective, i)
}
if utils.IntInSlice(i, Followers) == true {
Corrispective = append(Corrispective, i)
}
}
fmt.Println("\tFollowers: " + strconv.Itoa(len(Followers)))
fmt.Println("\tFollowing: " + strconv.Itoa(len(Following)))
fmt.Println("\tFollowers && Following: " + strconv.Itoa(len(Corrispective)))
fmt.Println("\tBetween mentions, those are whom the user is following: " + strconv.Itoa(len(MentionsWithCorrispective)))
//di := DemocracyIndex(len(myUniqueMentions), len(MentionsWithCorrispective), len(myTweets[i]), retweets)
om := OutsideMentions(len(myUniqueMentions), len(MentionsWithCorrispective))
apt := AnswerPeopleTax(len(myUniqueMentions), len(MentionsWithCorrispective), len(myTweets[i]), retweets)
// fmt.Println("\tDemocracy tax: " + FloatToString(di))
fmt.Println("\tOutside of circle mentions: " + FloatToString(om))
fmt.Println("\t of answering to external people: " + FloatToString(apt))
db.Create(i, "followers", []byte(strconv.Itoa(len(Followers))))
db.Create(i, "following", []byte(strconv.Itoa(len(Following))))
db.Create(i, "followers_followed", []byte(strconv.Itoa(len(Corrispective))))
db.Create(i, "mentions_to_followed", []byte(strconv.Itoa(len(MentionsWithCorrispective))))
}
fmt.Println("\tof wich, there are " + strconv.Itoa(retweets) + " retweets")
fmt.Println("\tof wich, there are " + strconv.Itoa(len(myUniqueMentions)) + " unique mentions (not in retweets)")
fmt.Println("\tof wich, there are " + strconv.Itoa(mymentions) + " total mentions (not in retweets)")
db.Create(i, "from", []byte(conf.Date))
db.Create(i, "tweets", []byte(strconv.Itoa(len(myTweets[i]))))
db.Create(i, "retweets", []byte(strconv.Itoa(retweets)))
db.Create(i, "unique_mentions", []byte(strconv.Itoa(len(myUniqueMentions))))
db.Create(i, "total_mentions", []byte(strconv.Itoa(mymentions)))
for k, v := range myUniqueMentions {
db.Create(i, strconv.FormatInt(k, 10), []byte(strconv.Itoa(v)), "map_unique_mentions")
}
// Visualize example:
//http://bl.ocks.org/mbostock/4062045
// Circle size is defined by it's radius (r) : .attr("r", 5)
// TOOlTIP: http://bl.ocks.org/Caged/6476579
}
}
func GatherDataFromAccount(crawler *TwitterCrawler, account string, timeLine timelinesTweets) {
retweetRegex, _ := regexp.Compile(`^RT`) // detecting retweets
log.Info(">> Depth look on " + account)
retweets := 0
db := nutz.NewStorage(account+".db", 0600, nil)
fmt.Println("-== Account: " + account + " ==-")
fmt.Println("\tTweets: " + strconv.Itoa(len(timeLine)))
var SocialNetwork map[string]struct{}
SocialNetwork = make(map[string]struct{})
for _, t := range timeLine {
// detecting hashtags
for _, tag := range t.Entities.Hashtags {
if tag.Text != "" {
fmt.Println("\tFound hashtag: " + tag.Text)
SocialNetwork[tag.Text] = struct{}{}
}
}
if retweetRegex.MatchString(t.Text) == true {
retweets++
}
}
fmt.Println("\tRetweets " + strconv.Itoa(retweets) + " retweets")
fmt.Println("\t" + strconv.Itoa(len(SocialNetwork)) + " hashtags")
var memory_network map[string]map[string]int
memory_network = make(map[string]map[string]int)
//Cycling on hashtags
for k, _ := range SocialNetwork {
db.Create(account, k, []byte(""), GENERATED_HASHTAGS)
db.Create(account, "retweets", []byte(strconv.Itoa(retweets)))
fmt.Println("\t Searching hashtag: " + k)
var MyTweetsNetwork searchTweets
// not searching right before we found an hashtag
// storing them to be UNIQUE, then in another phase searching deep further
if crawler.configuration.Number != 0 {
MyTweetsNetwork = crawler.SearchN("#"+k, crawler.configuration.Number, crawler.configuration.Slices)
} else {
MyTweetsNetwork = crawler.Search("#" + k)
}
for _, tweet := range MyTweetsNetwork {
if _, exists := memory_network[tweet.User.IdStr]; exists {
memory_network[tweet.User.ScreenName][k]++
} else {
memory_network[tweet.User.ScreenName] = make(map[string]int)
memory_network[tweet.User.ScreenName][k]++
}
}
}
for user, tags := range memory_network {
for tag, occurrence := range tags {
db.Create(account, tag, []byte(
strconv.Itoa(occurrence)), MATCHING_HASHTAGS, user)
}
}
}