How can I easily compare the distributions of multiple cohorts?
Usually, https://seaborn.pydata.org/generated/seaborn.distplot.html would be a great tool to visually compare distributions. However, due to the size of my dataset, I needed to compress it and only keep the counts.
It was created as:
SELECT age, gender, compress_distributionUDF(collect_list(struct(target_y_n, count, distribution_value))) GROUP BY age, gender
where compress_distributionUDF simply takes a list of tuples and returns the counts per group.
This leaves me with a list of
Row(distribution_value=60.0, count=314251, target_y_n=0)
nested inside a pandas.Series, but one per each chohort.
Basically, it is similar to:
pd.DataFrame({'foo':[1,2], 'bar':['first', 'second'], 'baz':[{'target_y_n': 0, 'value': 0.5, 'count':1000},{'target_y_n': 1, 'value': 1, 'count':10000}]})
and I wonder how to compare distributions:
within a cohort 0 vs. 1 of target_y_n
over multiple cohorts
in a way which is visually still understandable and not only a mess.
edit
For a single cohort Plotting pre aggregated data in python could be the answer, but how can multiple cohorts be compared (not just in a loop) as this leads to too many plots to compare?
I am still quite confused but we can start from this and see where it goes. From your example, I am focusing on baz as it is not clear to me what foo and bar are (I assume cohorts).
So let focus on baz and plot the different distributions according to target_y_n.
sns.catplot('value','count',data=df, kind='bar',hue='target_y_n',dodge=False,ci=None)
sns.catplot('value','count',data=df, kind='box',hue='target_y_n',dodge=False)
plt.bar(df[df['target_y_n']==0]['value'],df[df['target_y_n']==0]['count'],width=1)
plt.bar(df[df['target_y_n']==1]['value'],df[df['target_y_n']==1]['count'],width=1)
plt.legend(['Target=0','Target=1'])
sns.barplot('value','count',data=df, hue = 'target_y_n',dodge=False,ci=None)
Finally try to have a look at the FacetGrid class to extend your comparison (see here).
g=sns.FacetGrid(df,col='target_y_n',hue = 'target_y_n')
g=g.map(sns.barplot,'value','count',ci=None)
In your case you would have something like:
g=sns.FacetGrid(df,col='target_y_n',row='cohort',hue = 'target_y_n')
g=g.map(sns.barplot,'value','count',ci=None)
And a qqplot option:
from scipy import stats
def qqplot(x, y, **kwargs):
_, xr = stats.probplot(x, fit=False)
_, yr = stats.probplot(y, fit=False)
plt.scatter(xr, yr, **kwargs)
g=sns.FacetGrid(df,col='cohort',hue = 'target_y_n')
g=g.map(qqplot,'value','count')
Related
I have a datafile with quality scores from different suppliers over a time range of 3 years. The end goal is to use machine learning to predict the quality label (good or bad) of a shipment based on supplier information.
I want to use the mean historic quality data over a specific period of time as an input feature in this model by using pandas rolling window. the problem with this method is that pandas only allows you to create a window from t=0-x until t=0 for you rolling window as presented below:
df['average_score t-2w'] = df['score'].rolling(window='14d',closed='left').mean()
And this is were the problem comes in. For my feature I want to use quality data from a period of 2 weeks, but these 2 weeks are not the 2 weeks before the corresponding shipment, but of 2 weeks, starting from t=-4weeks , and ending on t=-2weeks.
You would imagine that this could be solved by using the same string of code but changing the window as presented below:
df['average_score t-2w'] = df['score'].rolling(window='28d' - '14d',closed='left').mean()
This, or any other type of denotation of this specific window does not seem to work.
It seems like pandas does not offer a solution to this problem, so we made a work around it with the following solution:
def time_shift_week(df):
def _avg_score_interval_func(series):
current_time = series.index[-1]
result = series[(series.index > ( current_time- pd.Timedelta(value=4, unit='w')))
& (series.index < (current_time - pd.Timedelta(value=2, unit='w')))]
return result.mean() if len(result)>0 else 0.0
temp_df = df.groupby(by=["supplier", "timestamp"], as_index=False).aggregate({"score": np.mean}).set_index('timestamp')
temp_df["w-42"] = (
temp_df
.groupby(["supplier"])
.ag_score
.apply(lambda x:
x
.rolling(window='30D', closed='both')
.apply(_avg_score_interval_func)
))
return temp_df.reset_index()
This results in a new df in which we find the average score score per supplier per timestamp, which we can subsequently merge with the original data frame to obtain the new feature.
Doing it this way seems really cumbersome and overly complicated for the task I am trying to perform. Eventhough we have found a workaround, I am wondering if there is an easier method of doing this.
Is anyone aware of a less complicated way of performing this rolling window feature extraction?
While pandas does not have the custom date offset you need, calculating the mean is pretty simple: it's just sum divided by count. You can subtract the 14-day rolling window from the 28-day rolling window:
# Some sample data. All scores are sequential for easy verification
idx = pd.MultiIndex.from_product(
[list("ABC"), pd.date_range("2020-01-01", "2022-12-31")],
names=["supplier", "timestamp"],
)
df = pd.DataFrame({"score": np.arange(len(idx))}, index=idx).reset_index()
# Now we gonna do rolling avg on score with the custom window.
# closed=left mean the current row will be excluded from the window.
score = df.set_index("timestamp").groupby("supplier")["score"]
r28 = score.rolling("28d", closed="left")
r14 = score.rolling("14d", closed="left")
avg_score = (r28.sum() - r14.sum()) / (r28.count() - r14.count())
I have a model that predicts 10 words for a particular course in order of likelihood, and I'd like the first 5 words of those words that appear in the course's description.
This is the format of the data:
course_name course_title course_description predicted_word_10 predicted_word_9 predicted_word_8 predicted_word_7 predicted_word_6 predicted_word_5 predicted_word_4 predicted_word_3 predicted_word_2 predicted_word_1
Xmath 32 Precalculus Polynomial and rational functions, exponential... directed scholars approach build african different visual cultures placed global
Xphilos 2 Morality Introduction to ethical and political philosop... make presentation weekly european ways general range questions liberal speakers
My idea is for each row to start iterating from predicted_word_1 until I get the first 5 that are in the description. I'd like to save those words in the order they appear into additional columns description_word_1 ... description_word_5. (If there are <5 predicted words in the description I plan to return NAN in the corresponding columns).
To clarify with an example: if the course_description of a course is 'Polynomial and rational functions, exponential and logarithmic functions, trigonometry and trigonometric functions. Complex numbers, fundamental theorem of algebra, mathematical induction, binomial theorem, series, and sequences. ' and its first few predicted words are irrelevantword1, induction, exponential, logarithmic, irrelevantword2, polynomial, algebra...
I would want to return induction, exponential, logarithmic, polynomial, algebra for that in that order and do the same for the rest of the courses.
My attempt was to define an apply function that will take in a row and iterate from the first predicted word until it finds the first 5 that are in the description, but the part I am unable to figure out is how to create these additional columns that have the correct words for each course. This code will currently only keep the words for one course for all the rows.
def find_top_description_words(row):
print(row['course_title'])
description_words_index=1
for i in range(num_words_per_course):
description = row.loc['course_description']
word_i = row.loc['predicted_word_' + str(i+1)]
if (word_i in description) & (description_words_index <=5) :
print(description_words_index)
row['description_word_' + str(description_words_index)] = word_i
description_words_index += 1
df.apply(find_top_description_words,axis=1)
The end goal of this data manipulation is to keep the top 10 predicted words from the model and the top 5 predicted words in the description so the dataframe would look like:
course_name course_title course_description top_description_word_1 ... top_description_word_5 predicted_word_1 ... predicted_word_10
Any pointers would be appreciated. Thank you!
If I understand correctly:
Create new DataFrame with just 100 predicted words:
pred_words_lists = df.apply(lambda x: list(x[3:].dropna())[::-1], axis = 1)
Please note that, there are lists in each row with predicted words. The order is nice, I mean the first, not empty, predicted word is on the first place, the second on the second place and so on.
Now let's create a new DataFrame:
pred_words_df = pd.DataFrame(pred_words_lists.tolist())
pred_words_df.columns = df.columns[:2:-1]
And The final DataFrame:
final_df = df[['course_name', 'course_title', 'course_description']].join(pred_words_df.iloc[:,0:11])
Hope this works.
EDIT
def common_elements(xx, yy):
temp = pd.Series(range(0, len(xx)), index= xx)
return list(df.reindex(yy).sort_values()[0:10].dropna().index)
pred_words_lists = df.apply(lambda x: common_elements(x[2].replace(',','').split(), list(x[3:].dropna())), axis = 1)
Does it satisfy your requirements?
Adapted solution (OP):
def get_sorted_descriptions_words(course_description, predicted_words, k):
description_words = course_description.replace(',','').split()
predicted_words_list = list(predicted_words)
predicted_words = pd.Series(range(0, len(predicted_words_list)), index=predicted_words_list)
predicted_words = predicted_words[~predicted_words.index.duplicated()]
ordered_description = predicted_words.reindex(description_words).dropna().sort_values()
ordered_description_list = pd.Series(ordered_description.index).unique()[:k]
return ordered_description_list
df.apply(lambda x: get_sorted_descriptions_words(x['course_description'], x.filter(regex=r'predicted_word_.*'), k), axis=1)
let's say i have a dataframe of 80 columns and 1 target column,
for example a bank account table with 80 attributes for each record (account) and 1 target column which decides if the client stays or leaves.
what steps and algorithms should i follow to select the most effective columns with the higher impact on the target column ?
There are a number of steps you can take, I'll give some examples to get you started:
A correlation coefficient, such as Pearson's Rho (for parametric data) or Spearman's R (for ordinate data).
Feature importances. I like XGBoost for this, as it includes the handy xgb.ggplot.importance / xgb.plot_importance methods.
One of the many feature selection options, such as python's sklearn.feature_selection methods.
This one way to do it using the Pearson correlation coefficient in Rstudio, I used it once when exploring the red_wine dataset my targeted variable or column was the quality and I wanted to know the effect of the rest of the columns on it.
see below figure shows the output of the code as you can see the blue color represents positive relation and red represents negative relations and the closer the value to 1 or -1 the darker the color
c <- cor(
red_wine %>%
# first we remove unwanted columns
dplyr::select(-X) %>%
dplyr::select(-rating) %>%
mutate(
# now we translate quality to a number
quality = as.numeric(quality)
)
)
corrplot(c, method = "color", type = "lower", addCoef.col = "gray", title = "Red Wine Variables Correlations", mar=c(0,0,1,0), tl.cex = 0.7, tl.col = "black", number.cex = 0.9)
I've used NLTK to pos_tag sentences in a pandas dataframe from an old Yelp competition. This returns a list of tuples (word, POS). I'd like to count the number of parts of speech for each instance. How would I, say, create a function to count the number of being verbs in each review? I know how to apply functions to features - no problem there. I just can't wrap my head around how to count things inside tuples inside lists inside a pd feature.
The head is here, as a tsv: https://pastebin.com/FnnBq9rf
Thank you #zhangyulin for your help. After two days, I learned some incredibly important things (as a novice programmer!). Here's the solution!
def NounCounter(x):
nouns = []
for (word, pos) in x:
if pos.startswith("NN"):
nouns.append(word)
return nouns
df["nouns"] = df["pos_tag"].apply(NounCounter)
df["noun_count"] = df["nouns"].str.len()
As an example, for dataframe df, noun count of the column "reviews" can be saved to a new column "noun_count" using this code.
def NounCount(x):
nounCount = sum(1 for word, pos in pos_tag(word_tokenize(x)) if pos.startswith('NN'))
return nounCount
df["noun_count"] = df["reviews"].apply(NounCount)
df.to_csv('./dataset.csv')
There are a number of ways you can do that and one very straight forward way is to map the list (or pandas series) of tuples to indicator of whether the word is a verb, and count the number of 1's you have.
Assume you have something like this (please correct me if it's not, as you didn't provide an example):
a = pd.Series([("run", "verb"), ("apple", "noun"), ("play", "verb")])
You can do something like this to map the Series and sum the count:
a.map(lambda x: 1 if x[1]== "verb" else 0).sum()
This will return you 2.
I grabbed a sentence from the link you shared:
text = nltk.word_tokenize("My wife took me here on my birthday for breakfast and it was excellent.")
tag = nltk.pos_tag(text)
a = pd.Series(tag)
a.map(lambda x: 1 if x[1]== "VBD" else 0).sum()
# this returns 2
I'm very confused on the apply function for pandas. I have a big dataframe where one column is a column of strings. I'm then using a function to count part-of-speech occurrences. I'm just not sure the way of setting up my apply statement or my function.
def noun_count(row):
x = tagger(df['string'][row].split())
# array flattening and filtering out all but nouns, then summing them
return num
So basically I have a function similar to the above where I use a POS tagger on a column that outputs a single number (number of nouns). I may possibly rewrite it to output multiple numbers for different parts of speech, but I can't wrap my head around apply.
I'm pretty sure I don't really have either part arranged correctly. For instance, I can run noun_count[row] and get the correct value for any index but I can't figure out how to make it work with apply how I have it set up. Basically I don't know how to pass the row value to the function within the apply statement.
df['num_nouns'] = df.apply(noun_count(??),1)
Sorry this question is all over the place. So what can I do to get a simple result like
string num_nouns
0 'cat' 1
1 'two cats' 1
EDIT:
So I've managed to get something working by using list comprehension (someone posted an answer, but they've deleted it).
df['string'].apply(lambda row: noun_count(row),1)
which required an adjustment to my function:
def tagger_nouns(x):
list_of_lists = st.tag(x.split())
flat = [y for z in list_of_lists for y in z]
Parts_of_speech = [row[1] for row in flattened]
c = Counter(Parts_of_speech)
nouns = c['NN']+c['NNS']+c['NNP']+c['NNPS']
return nouns
I'm using the Stanford tagger, but I have a big problem with computation time, and I'm using the left 3 words model. I'm noticing that it's calling the .jar file again and again (java keeps opening and closing in the task manager) and maybe that's unavoidable, but it's really taking far too long to run. Any way I can speed it up?
I don't know what 'tagger' is but here's a simple example with a word count that ought to work more or less the same way:
f = lambda x: len(x.split())
df['num_words'] = df['string'].apply(f)
string num_words
0 'cat' 1
1 'two cats' 2