Cleaning/Basic Exploration

import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
import math

rawData = pd.read_csv("renamedHeaders.csv", skiprows=[0])

rawData.head()

	Start Date	End Date	IP Address	Progress	Duration (in seconds)	Finished	Recorded Date	Response ID	Recipient Last Name	Recipient First Name	...	External Data Reference	Location Latitude	Location Longitude	Distribution Channel	User Language	campus	educationLevel	isGrad	experience	continued
0	3/25/21 12:03	3/25/21 12:05	67.183.121.24	100	95	True	3/25/21 12:05	R_YQ7iSdzAhzSuTu1	NaN	NaN	...	NaN	47.497299	-122.355499	anonymous	EN	Seattle	Undergraduate student	0	Mostly negative	My professors have done a great job overall wi...
1	3/25/21 12:04	3/25/21 12:05	67.170.122.210	100	39	True	3/25/21 12:05	R_3jGNkoQPznVw0Df	NaN	NaN	...	NaN	47.749496	-122.297600	anonymous	EN	Seattle	Undergraduate student	0	A mix of positive and negative	Also providing online resources like updating ...
2	3/25/21 12:07	3/25/21 12:07	98.232.33.143	100	17	True	3/25/21 12:07	R_9ZDweGffIIsCve1	NaN	NaN	...	NaN	47.651703	-117.083801	anonymous	EN	Seattle	Undergraduate student	0	Mostly positive	NaN
3	3/25/21 12:05	3/25/21 12:07	172.58.44.174	100	120	True	3/25/21 12:07	R_29hsSW6x49hZS6c	NaN	NaN	...	NaN	47.900497	-122.247200	anonymous	EN	Bothell	Undergraduate student	0	A mix of positive and negative	I enjoy the zoom session options, and while I ...
4	3/25/21 12:06	3/25/21 12:07	73.254.146.138	100	57	True	3/25/21 12:07	R_1I61O4XyIcHW8A3	NaN	NaN	...	NaN	47.698700	-117.439697	anonymous	EN	Seattle	Undergraduate student	0	A mix of positive and negative	Flexible class time, less waste of time, less ...

5 rows × 21 columns

rawData.shape

(3769, 21)

rawData.isna().sum()

Start Date                    0
End Date                      0
IP Address                    0
Progress                      0
Duration (in seconds)         0
Finished                      0
Recorded Date                 0
Response ID                   0
Recipient Last Name        3769
Recipient First Name       3769
Recipient Email            3769
External Data Reference    3769
Location Latitude             0
Location Longitude            0
Distribution Channel          0
User Language                 0
campus                        2
educationLevel                2
isGrad                        0
experience                    4
continued                   494
dtype: int64

campusTotals = nonEmptyC.groupby("campus").ones.sum()
campusTotals

campus
Bothell     864.0
Seattle    2110.0
Tacoma      297.0
Name: ones, dtype: float64

nonEmptyC.groupby("educationLevel").ones.sum()

educationLevel
Graduate student          634.0
Undergraduate student    2637.0
Name: ones, dtype: float64

nonEmptyExp = rawData[~rawData.experience.isna()].copy()

### Adding a numerical positive/negative experience
def getExpInt(text):
    if ("Mostly positive" in text):
        return 1
    elif ("Mostly negative" in text):
        return -1
    else:
        return 0
nonEmptyExp["experienceInt"] = nonEmptyExp.apply(lambda x: getExpInt(x.experience), axis=1)
nonEmptyExp[["experienceInt", "experience"]].head(3)

	experienceInt	experience
0	-1	Mostly negative
1	0	A mix of positive and negative
2	1	Mostly positive

Experiences across the sample

nonEmptyExp.groupby("educationLevel").experienceInt.mean()

educationLevel
Graduate student         0.198087
Undergraduate student    0.094688
Name: experienceInt, dtype: float64

nonEmptyExp.groupby("campus").experienceInt.mean()

campus
Bothell    0.263158
Seattle    0.032862
Tacoma     0.250000
Name: experienceInt, dtype: float64

nonEmptyC = nonEmptyExp[~nonEmptyExp.continued.isna()].copy()

nonEmptyC.isna().sum()

Start Date                    0
End Date                      0
IP Address                    0
Progress                      0
Duration (in seconds)         0
Finished                      0
Recorded Date                 0
Response ID                   0
Recipient Last Name        3273
Recipient First Name       3273
Recipient Email            3273
External Data Reference    3273
Location Latitude             0
Location Longitude            0
Distribution Channel          0
User Language                 0
campus                        2
educationLevel                2
isGrad                        0
experience                    0
continued                     0
experienceInt                 0
dtype: int64

Isolating Mentions of Commuting

nonEmptyC["commute"] = nonEmptyC.apply(lambda x: int("commut" in x.continued), axis=1)

nonEmptyC.loc[nonEmptyC.commute == 1].shape

(143, 25)

round(nonEmptyC.groupby("commute").experienceInt.mean(), 2)

commute
0    0.11
1    0.50
Name: experienceInt, dtype: float64

Isolating Mentions of Recordings

nonEmptyC["record"] = nonEmptyC.apply(lambda x: int("record" in x.continued.lower()), axis=1)

nonEmptyC.record.sum()

938

nonEmptyC.shape[0]

3273

round(938/3273, 2)

0.29

Isolating Office Hours

nonEmptyC["officeHrs"] = nonEmptyC.apply(lambda x: containsSynonym(x.continued, ["office hours", "office-hours"]), axis=1)

nonEmptyC.officeHrs.sum()

264

gradOfficeHrs = nonEmptyC.groupby("isGrad").officeHrs.sum()
gradOfficeHrs

isGrad
0    222
1     42
Name: officeHrs, dtype: int64

gradTotals = nonEmptyC.groupby("isGrad").ones.sum()
gradTotals

isGrad
0    2637.0
1     636.0
Name: ones, dtype: float64

round(gradTotals / gradOfficeHrs, 1)

isGrad
0    11.9
1    15.1
dtype: float64

Isolating Open-Book synonyms

openBookSyn = ["open book", "open-book", "open note", "open-note", "take home", "take-home"]

def containsSynonym(text, synonyms): # Checks if given text contains any of the given synonyms
    text = text.lower()
    for token in synonyms:
        if token in text:
            return 1
    return 0

nonEmptyC["openBook"] = nonEmptyC.apply(lambda x: containsSynonym(x.continued, openBookSyn), axis=1)

nonEmptyC.openBook.sum()

93

nonEmptyC["exam"] = nonEmptyC.apply(lambda x: containsSynonym(x.continued, ["quiz", "exam"]), axis=1)

nonEmptyC.exam.sum()

255

nonEmptyC.loc[(nonEmptyC.exam == 1) & (nonEmptyC.openBook == 1)].shape

(61, 31)

round(61/255, 2)

0.24

255-61

194

# Generating data for honeycomb visual
bookMentions = np.append(np.ones(shape=61), np.zeros(shape=(255-61)))
np.random.shuffle(bookMentions)
np.savetxt("openBook.csv", bookMentions, delimiter=",")

nonEmptyC["nada"] = nonEmptyC.apply(lambda x: containsSynonym(x.continued, ["nothing", "none", "n/a"]), axis=1)
nonEmptyC.nada.sum()

189

Asynchronous mentions

nonEmptyC["async"] = nonEmptyC.apply(lambda x: containsSynonym(x.continued, ["asynchronous", "hybrid"]), axis=1)

nonEmptyC["async"].sum()

468

round(468/3273, 2)

0.14

Professors and leniency

softies = ["lenien", "understanding", "grace", "flexibility"]

nonEmptyC["includedProf"] = nonEmptyC.apply(lambda x: "prof" in x.continued.lower(), axis=1)
nonEmptyC["softerProfs"] = nonEmptyC.apply(lambda x: containsSynonym(x.continued, softies), axis=1)

nonEmptyC.includedProf.sum()

351

nonEmptyC.softerProfs.sum()

429

nonEmptyC.loc[(nonEmptyC.includedProf == True) & (nonEmptyC.softerProfs == True)].shape

(104, 30)

round(104/351, 2)

0.3

# Generating datapoints to use for scatter visualization
profs = np.append(np.ones(shape=104), np.zeros(shape=(351-104)))
np.random.shuffle(profs)
np.savetxt("softProf2.csv", profs, delimiter=",")

Three main categories - counts

flexList = ['save', 'saving', 'saves', 'commute', 'commuting', 'available', 'availability', 'conflict', 'freedom', 'manage', 'managing', 'flexibility', 'flexible', 'responsiveness', 'responsive', 'accommodation', 'able', 'own', 'allow', 'allowed', 'allows', 'plan', 'ahead', 'comfort', 'prepare', 'ability', 'choose', 'sick', 'parent', 'parents', 'balance', 'work', 'convenient', 'opens', 'leniency']

featuresList = ['S/NS', 'accommodation', 'grades', 'grade', 'grading', 'lecture', 'presentation', 'professors', 'professor', 'participation', 'record', 'recorded', 'pre-recorded', 'discussion', 'participant', 'powerpoint', 'canvas', 'extensions', 'understanding', 'open book', 'exams', 'exam', 'finals', 'midterms', 'final', 'midterm', 'take-home', 'note', 'test', 'policies', 'office hours', 'office hour', 'online resources', 'zoom',  'online communities', 'attendance', 'remote', 'group projects']

accessList = ['accessible', 'access', 'accessibility', 'option', 'accommodation', 'immunocompromised', 'disabled', 'disability', 'disabilities', 'struggles', 'help', 'anxious', 'difficult', 'subtitles', 'feasible', 'international', 'choice', 'concern', 'safe', 'resources', 'services', 'equipment', 'need', 'possible']

negsList = ['lack', 'not helpful', 'suffer', 'suffered', 'little', 'degraded', 'hate', 'uncomfortable', 'terrible', 'failed', 'fail', 'failing', 'hard to', 'negative', 'anxiety', 'procrastination', 'barely', 'not getting', 'joke', 'lost', 'nothing', 'none', 'n/a', 'back']

nonEmptyC["flexCt"] = nonEmptyC.apply(lambda x: containsSynonym(x.continued, flexList), axis=1)
nonEmptyC["featuresCt"] = nonEmptyC.apply(lambda x: containsSynonym(x.continued, featuresList), axis=1)
nonEmptyC["accessCt"] = nonEmptyC.apply(lambda x: containsSynonym(x.continued, accessList), axis=1)
nonEmptyC["negCt"] = nonEmptyC.apply(lambda x: containsSynonym(x.continued, negsList), axis=1)

# all responses
nonEmptyC.shape

(3273, 37)

nonEmptyC[["flexCt", "featuresCt", "accessCt", "negCt"]].sum()

flexCt        1589
featuresCt    2382
accessCt      1488
negCt          527
dtype: int64

nonEmptyC.groupby("campus")[["flexCt", "featuresCt", "accessCt", "negCt"]].sum()

	flexCt	featuresCt	accessCt	negCt
campus
Bothell	379	569	372	122
Seattle	1050	1619	976	361
Tacoma	158	193	139	44

Explicit Defintions

shortFlex = ['flexibil', 'asynchronous', 'hybrid']
shortAccess = ['record', 'access', 'commut']
shortFeatures = ['quiz', 'assignment', 'exam']

# Course Guidelines - open_note, take_home

def addCt(dictionary):
    for i in dictionary:
        name = i + "_ct"
        nonEmptyC[name] = nonEmptyC.apply(lambda x: containsSynonym(x.continued, [i]), axis=1)
addCt(shortFlex)
addCt(shortAccess)
addCt(shortFeatures)

nonEmptyC[["flexibil_ct", "asynchronous_ct", "hybrid_ct"]].sum()

flexibil_ct        308
asynchronous_ct    341
hybrid_ct          140
dtype: int64

campusFlex = nonEmptyC.groupby("campus")[["flexibil_ct", "asynchronous_ct", "hybrid_ct"]].sum()
campusFlex

	flexibil_ct	asynchronous_ct	hybrid_ct
campus
Bothell	62	61	66
Seattle	221	250	64
Tacoma	25	30	9

campusFlex["totals"] = campusFlex.flexibil_ct + campusFlex.asynchronous_ct + campusFlex.hybrid_ct

campusFlex.totals.sum()

788

campusTotals.T

campus
Bothell     864.0
Seattle    2110.0
Tacoma      297.0
Name: ones, dtype: float64

round(campusFlex / 788, 2)

	flexibil_ct	asynchronous_ct	hybrid_ct	totals
campus
Bothell	0.08	0.08	0.08	0.24
Seattle	0.28	0.32	0.08	0.68
Tacoma	0.03	0.04	0.01	0.08

campusFlex["campusTotal"] = campusTotals

campusFlex.iloc[0] / campusTotals[0]

flexibil_ct        0.071759
asynchronous_ct    0.070602
hybrid_ct          0.076389
totals             0.218750
campusTotal        1.000000
Name: Bothell, dtype: float64

campusFlex.iloc[1] / campusTotals[1]

flexibil_ct        0.104739
asynchronous_ct    0.118483
hybrid_ct          0.030332
totals             0.253555
campusTotal        1.000000
Name: Seattle, dtype: float64

campusFlex.iloc[2] / campusTotals[2]

flexibil_ct        0.084175
asynchronous_ct    0.101010
hybrid_ct          0.030303
totals             0.215488
campusTotal        1.000000
Name: Tacoma, dtype: float64

normalizedFlex

flexibil_ct                                                0.0717593
asynchronous_ct                                            0.0706019
hybrid_ct                                                  0.0763889
totals                                                       0.21875
bothell            flexibil_ct        0.071759
asynchronous_ct   ...
seattle            flexibil_ct        0.104739
asynchronous_ct   ...
tacoma             flexibil_ct        0.084175
asynchronous_ct   ...
Name: Bothell, dtype: object

nonEmptyC[["quiz_ct", "exam_ct", "assignment_ct"]].sum()

quiz_ct           82
exam_ct          201
assignment_ct    168
dtype: int64

campusFeatures = nonEmptyC.groupby("campus")[["quiz_ct", "exam_ct", "assignment_ct"]].sum()
campusFeatures["cTotal"] = campusTotals
campusFeatures

	quiz_ct	exam_ct	assignment_ct	cTotal
campus
Bothell	10	38	39	864.0
Seattle	67	153	112	2110.0
Tacoma	4	10	17	297.0

campusFeatures["normQuiz"] = campusFeatures.quiz_ct / campusFeatures.cTotal
campusFeatures["normAss"] = campusFeatures.assignment_ct / campusFeatures.cTotal
campusFeatures["normExam"] = campusFeatures.exam_ct / campusFeatures.cTotal

campusFeatures[["normQuiz", "normAss", "normExam"]]

	normQuiz	normAss	normExam
campus
Bothell	0.011574	0.045139	0.043981
Seattle	0.031754	0.053081	0.072512
Tacoma	0.013468	0.057239	0.033670

campusTotals

campus
Bothell     864.0
Seattle    2110.0
Tacoma      297.0
Name: ones, dtype: float64

nonEmptyC[["record_ct", "access_ct", "commut_ct"]].sum()

record_ct    938
access_ct    289
commut_ct    144
dtype: int64

campusAccess = nonEmptyC.groupby("campus")[["record_ct", "access_ct", "commut_ct"]].sum()
campusAccess

	record_ct	access_ct	commut_ct
campus
Bothell	161	43	43
Seattle	720	229	84
Tacoma	57	16	17

campusAccess["cTotal"] = campusTotals

campusAccess["recordNorm"] = campusAccess.record_ct / campusAccess.cTotal
campusAccess["accessNorm"] = campusAccess.access_ct / campusAccess.cTotal
campusAccess["commutdNorm"] = campusAccess.commut_ct / campusAccess.cTotal
campusAccess[["commutdNorm", "recordNorm", "accessNorm"]]

	commutdNorm	recordNorm	accessNorm
campus
Bothell	0.049769	0.186343	0.049769
Seattle	0.039810	0.341232	0.108531
Tacoma	0.057239	0.191919	0.053872

nonEmptyC["sFlexCt"] = nonEmptyC.apply(lambda x: containsSynonym(x.continued, shortFlex), axis=1)
nonEmptyC["sFeaturesCt"] = nonEmptyC.apply(lambda x: containsSynonym(x.continued, shortAccess), axis=1)
nonEmptyC["sAccessCt"] = nonEmptyC.apply(lambda x: containsSynonym(x.continued, shortFeatures), axis=1)

nonEmptyC[["sFlexCt", "sFeaturesCt", "sAccessCt"]].sum()

sFlexCt         717
sFeaturesCt    1202
sAccessCt       394
dtype: int64

nonEmptyC.groupby("campus")[["sFlexCt", "sFeaturesCt", "sAccessCt"]].sum()

	sFlexCt	sFeaturesCt	sAccessCt
campus
Bothell	179	224	76
Seattle	479	894	289
Tacoma	58	83	28

Some NLP to explore the open-text question

### importing nltk stopwords broken; hardcoded temporarily
stopWords = ['i', 'me', 'my', 'myself', 'we', 'our', 'ours', 'ourselves', 'you', "you're", "you've", "you'll", "you'd", 'your', 'yours', 'yourself', 'yourselves', 'he', 'him', 'his', 'himself', 'she', "she's", 'her', 'hers', 'herself', 'it', "it's", 'its', 'itself', 'they', 'them', 'their', 'theirs', 'themselves', 'what', 'which', 'who', 'whom', 'this', 'that', "that'll", 'these', 'those', 'am', 'is', 'are', 'was', 'were', 'be', 'been', 'being', 'have', 'has', 'had', 'having', 'do', 'does', 'did', 'doing', 'a', 'an', 'the', 'and', 'but', 'if', 'or', 'because', 'as', 'until', 'while', 'of', 'at', 'by', 'for', 'with', 'about', 'against', 'between', 'into', 'through', 'during', 'before', 'after', 'above', 'below', 'to', 'from', 'up', 'down', 'in', 'out', 'on', 'off', 'over', 'under', 'again', 'further', 'then', 'once', 'here', 'there', 'when', 'where', 'why', 'how', 'all', 'any', 'both', 'each', 'few', 'more', 'most', 'other', 'some', 'such', 'no', 'nor', 'not', 'only', 'own', 'same', 'so', 'than', 'too', 'very', 's', 't', 'can', 'will', 'just', 'don', "don't", 'should', "should've", 'now', 'd', 'll', 'm', 'o', 're', 've', 'y', 'ain', 'aren', "aren't", 'couldn', "couldn't", 'didn', "didn't", 'doesn', "doesn't", 'hadn', "hadn't", 'hasn', "hasn't", 'haven', "haven't", 'isn', "isn't", 'ma', 'mightn', "mightn't", 'mustn', "mustn't", 'needn', "needn't", 'shan', "shan't", 'shouldn', "shouldn't", 'wasn', "wasn't", 'weren', "weren't", 'won', "won't", 'wouldn', "wouldn't"]

wordCount = pd.DataFrame(columns = ["word", "ct", "posCt", "negCt", "ntCt"])

## Sweep through text and add frequency of single and double_words to appropriate dictionaries (overall counts and pos/neg counts)
def countWords(text, dictionary, exp): 
    # Choose appropriate experience group
    headerDict = ["negCt", "ntCt", "posCt"]
    ctHeader = headerDict[exp + 1]
    # Tokenize Text
    words = text.split(" ")
    prev_word = ""
    global wordCount
    # Add each token to the frequency dictionaries
    for curWord in words:
        curWord = curWord.lower().strip('!?(.,\')')
        # Add Single Token
        if ((wordCount.loc[wordCount.word == curWord].shape[0] == 0) and (not (curWord in stopWords))):  # Append new-showing words
            wordCount = wordCount.append({'word': curWord, 'ct': 0, 'posCt': 0, 'negCt':0, 'ntCt': 0}, ignore_index=True)
        wordCount.loc[wordCount.word == curWord, [dictionary, ctHeader]] += 1 # Increase Both Counts
        # Add Double_Tokens
        if ((prev_word != "") and (not ((prev_word in stopWords) and (curWord in stopWords)))):
            concat = prev_word + "_" + curWord
            if (wordCount.loc[wordCount.word == concat].shape[0] == 0):  # add new entry if new word
                wordCount = wordCount.append({'word': concat, 'ct': 0, 'posCt': 0, 'negCt':0, 'ntCt': 0}, ignore_index=True)
            wordCount.loc[wordCount.word == concat, [dictionary, ctHeader]] += 1
        prev_word = curWord # Reset for next itertion

nonEmptyC.apply(lambda x: countWords(x.continued, "ct", x.experienceInt), axis=1)

0       None
1       None
3       None
4       None
5       None
        ... 
3764    None
3765    None
3766    None
3767    None
3768    None
Length: 3273, dtype: object

wordCount.head()

	word	ct	posCt	negCt	ntCt
0	professors	338	68	52	218
1	my_professors	20	2	4	14
2	professors_have	25	4	2	19
3	done	46	10	12	24
4	have_done	10	0	3	7

nonEmptyC["ones"] = np.ones(shape=nonEmptyC.shape[0])
responses = nonEmptyC.groupby("experienceInt").ones.sum()
responses

experienceInt
-1     502.0
 0    1840.0
 1     931.0
Name: ones, dtype: float64

responseRatio = responses[-1]/responses[1]
responseRatio

0.5392051557465092

wordCount.loc[wordCount.posCt == 0, "posCt"] = 1
wordCount["negRatio"] = wordCount.negCt / (wordCount.posCt * responseRatio)

wordCount.sort_values(by="negRatio", ascending=False).head(10)

	word	ct	posCt	negCt	ntCt	negRatio
733	none	134	1	108	25	200.295
6334	none_of	30	1	28	2	51.9283
413	nothing	57	4	35	18	16.2276
1985	none_i	10	1	8	2	14.8367
8135	almost	21	2	14	5	12.9821
14008	struggle	9	1	7	1	12.9821
1993	tuition	22	2	13	7	12.0548
4010	paying_full	6	1	6	0	11.1275
4011	full_tuition	6	1	6	0	11.1275
3790	experience_i	10	1	6	3	11.1275

math.log(2)

0.6931471805599453

wordCount["logNegRatio"] = np.log(wordCount.negRatio.astype(float))

wordCount.sort_values(by="logNegRatio", ascending=False)

	word	ct	posCt	negCt	ntCt	negRatio	logNegRatio
733	none	134	1	108	25	200.295	5.299790
6334	none_of	30	1	28	2	51.9283	3.949864
413	nothing	57	4	35	18	16.2276	2.786713
1985	none_i	10	1	8	2	14.8367	2.697101
8135	almost	21	2	14	5	12.9821	2.563569
...	...	...	...	...	...	...	...
15759	this_continues	3	1	0	3	0	-inf
15758	and_efficient	3	1	0	2	0	-inf
15757	they've_been	2	1	0	2	0	-inf
15756	hours_they've	1	1	0	1	0	-inf
38952	meetings_will	1	1	0	0	0	-inf

38953 rows × 7 columns

wordCount.drop(wordCount.loc[wordCount.word == "none_i"].index, inplace=True)

freqDict = pd.Series(wordCount["ct"].values,index=wordCount.word).to_dict()
posDict = pd.Series(wordCount["posCt"].values,index=wordCount.word).to_dict()
negDict = pd.Series(wordCount["negCt"].values,index=wordCount.word).to_dict()
ratioDict = pd.Series(wordCount["logNegRatio"].values, index=wordCount.word).to_dict()
len(freqDict)

38952

from wordcloud import WordCloud

def showWordCloud(frequencies, clrmap, save="na"):
    wordcloud = WordCloud(width = 800, height = 800,
                    background_color ='white',
                    colormap=clrmap,
                    min_font_size = 10).generate_from_frequencies(frequencies)

    # plot the WordCloud image                       
    plt.figure(figsize = (8, 8), facecolor = None)
    plt.imshow(wordcloud)
    plt.axis("off")
    plt.tight_layout(pad = 0)
    if (save != "na"):
        plt.savefig(save)
    plt.show()

showWordCloud(freqDict, singleShadePurp, "singleShades/inclusive.png")

showWordCloud(freqDict, newcmp, "inclPur2.png")

showWordCloud(posDict, "ocean")

showWordCloud(negDict, "hot")

huge = ["class", "person", "would"]
for x in huge:
    del posDict[x]
    del negDict[x]
print(len(posDict))

38942

showWordCloud(posDict, singleShadePurp, "singleShades/positives.png")

showWordCloud(posDict, newcmp, "finalPos.png")

showWordCloud(negDict, "hot", "shortenedNeg.png")

del ratioDict["experience_i"]

from matplotlib import cm
from matplotlib.colors import ListedColormap,LinearSegmentedColormap

# modified hsv in 256 color class
purples = cm.get_cmap('Purples', 256)
# create new hsv colormaps with custom range
darkPruples = ListedColormap(purples(np.linspace(0.4, 1, 256)))
singleShadePurp = ListedColormap(purples(np.linspace(0.85, 0.95, 256)))

posPurp = ListedColormap(purples(np.linspace(0.4, 0.8, 256)))

bupu = cm.get_cmap('twilight_shifted', 256)
negPurp = ListedColormap(bupu(np.linspace(0.04, 0.14, 256)))

grayc = cm.get_cmap('gray', 256)
greys = ListedColormap(grayc(np.linspace(0.2, 0.5, 256)))
singleShadeGrey = ListedColormap(grayc(np.linspace(0.3, 0.4, 256)))

showWordCloud(ratioDict, singleShadeGrey, "singleShades/negatives.png")

showWordCloud(ratioDict, greys, "greyNegs2.png")