CODEWITHSUNDEEP
rggplot2
Brian Lee
Brian Lee

Reputation: 87

t bar graph with negative values

I am trying to achieve something similar to this with a dataframe i have but i can't get it to work for me.

My dataframe is as follows;

Sentiment   word    freq
Negative    obik    52
Negative    ride    7
Negative    just    6
Negative    bad     6
Negative    like    6
Negative    bike    5
Negative    cycl    5
Negative    home    5
Negative    park    5
Negative    get     4
Positive    obik    71
Positive    ride    26
Positive    free    13
Positive    just    11
Positive    cycl    9
Positive    start   8
Positive    come    7
Positive    park    7
Positive    healthi 7
Positive    bike    7

What i want is a graph that looks like the image below where words that have a positive sentiment appear above the x-axis with their respective frequency and word with a negative sentiment appear below the x-axis with their respective frequency. The y-axis is the frequency of the word graph

After many failed attempts, i am back with the simple code that does not achieve what i want;

ggplot(AsiaSentiment10,aes(word,freq,label=""))+
geom_bar(stat="identity",position="identity")

What do i need to do to make my graph?

Upvotes: 1

Views: 1373

Answers (1)

Gregor Thomas
Gregor Thomas

Reputation: 146050

Maybe this is what you are looking for?

df$sign_freq = df$freq * ifelse(df$Sentiment == "Positive", 1, -1)
df$word = reorder(df$word, df$sign_freq, sum)

ggplot(df, aes(x = word, y = sign_freq, fill = Sentiment)) +
    geom_bar(position = 'identity', stat = "identity")

enter image description here

Or this?

df_agg = aggregate(sign_freq ~ word, FUN = sum, data = df)

ggplot(df_agg, aes(x = word, y = sign_freq, fill = factor(sign(sign_freq)))) +
    geom_bar(position = 'identity', stat = "identity") +
    guides(fill = FALSE) +
    labs(y = "diff")

enter image description here

Upvotes: 4

Related Questions