Mapping wordcloud color to a value for sentiment analysis

So I'm looking to see if there is a way to map the color of a word cloud to a value, or maybe even overlap two word clouds (one positive and one negative list) with the end result being a dark color for negative sentiment and a bright color for a positive sentiment like in the picture only this is random.

I'm not sure how you would assign the value because from what I see you either paste text or text with a frequency value but maybe the latter of two maps overlapped?

I was able to change the color to greens here by copying code from StackOverflow and adjusting the values. As far as I can tell it's just randomly assigned. I pasted that code just below

def green_color_func(word, font_size, position,orientation,random_state=None, **kwargs):
    return("hsl(100,100%%, %d%%)" % np.random.randint(1,51))
wordCloud.recolor(color_func = green_color_func)

below is a sample of my simple code, for clarity I removed stopwords, font path etc, mask and b3 is a dictionary of hashtags and a count frequency...maybe somehow with a colorfunc or recolor(self[, random_state, color_func, …])?...

wordCloud = WordCloud(font_path = font_path,width=1000, height=800,max_words=100,
                      random_state=21, background_color = 'white',
                      prefer_horizontal=1).generate_from_frequencies(b3)




plt.figure(figsize=(30,15))
plt.imshow(wordCloud, interpolation = 'bilinear')
plt.axis('off')
plt.show() 

wordCloud.to_file("bt3.png")

I'm fairly new to coding and Im stumped by this so I appreciate any insight, thanks.

Solution

I know this was asked a while back - and may no longer be relevant.
However, if someone is looking to achieve the same thing, then this is the way to do it.
Assuming you have a

list of words (word),
their sentiment (score),
and their frequency (freq)

in a dataframe like this:

import pandas as pd

# Create a dataframe with the sentimental word list
df = pd.read_csv('https://raw.githubusercontent.com/text-machine-lab/sentimental/master/sentimental/word_list/afinn.csv')

# Generate the random frequency for each word, however, set the seed for reproducibility
rs = np.random.seed(42)
df['freq'] = np.random.randint(1, 100, df.shape[0])

print(df)

OUTPUT:

           word  score  freq
0       abandon     -2    52
1     abandoned     -2    93
2      abandons     -2    15
3      abducted     -2    72
4     abduction     -2    61
...         ...    ...   ...
2457      yucky     -2    77
2458      yummy      3    57
2459     zealot     -2    25
2460    zealots     -2     2
2461    zealous      2    64

You could create a word cloud with colours mapped to their sentiments using the following code:

Note:

positive words are green,
negative words are red,
and neutral words are grey.

from wordcloud import WordCloud
import matplotlib.pyplot as plt
import pandas as pd
import numpy as np

class SimpleGroupedColorFunc(object):
    def __init__(self, color_to_words, default_color):
        self.word_to_color = {word: color
                              for (color, words) in color_to_words.items()
                              for word in words}

        self.default_color = default_color

    def __call__(self, word, **kwargs):
        return self.word_to_color.get(word, self.default_color)



# Create a dataframe with the sentimental word list
df = pd.read_csv('https://raw.githubusercontent.com/text-machine-lab/sentimental/master/sentimental/word_list/afinn.csv')

# Generate the random frequency for each word, however, set the seed for reproducibility
rs = np.random.seed(42)
df['freq'] = np.random.randint(1, 100, df.shape[0])

# Generate a word-frequency dictionary to import to the word cloud
text_freq_dict = dict(zip(df.word, df.freq))

# Generate word cloud from word-frequency dictionary
wordcloud = WordCloud()
wordcloud.generate_from_frequencies(frequencies=text_freq_dict)

pos_words = []
neg_words = []

# Categorize the words based on sentiment score, assign to appropriate list 
for word, sentiment in zip(df['word'], df['score']):
    if sentiment > 1:
        pos_words.append(word)
    elif sentiment < -1:
        neg_words.append(word)


# Create a dictionary containing the colours to be used for each sentiment list
color_to_words = {
    # green colour for positive words
    '#00ff00': pos_words,
    # red for negative words
    'red': neg_words
}

# Neutral words will be grey (with sentiment between -1 and 1 inclusive)
default_color = 'grey'

# Create a color function with simple groupded tones
grouped_color_func = SimpleGroupedColorFunc(color_to_words, default_color)

# Apply our color function to the word cloud
wordcloud.recolor(color_func=grouped_color_func)

# Plot the word cloud
plt.figure()
plt.imshow(wordcloud, interpolation="bilinear")
plt.axis("off")
plt.show()

OUTPUT:

Please note that some parts of the code above were borrowed from the wordcloud package site.