A question that haunts me a little bit. Though it must be a common task to perform, I find it difficult to implement it easily in pandas
you have a df_ex of values of category and score. based on the score value, you want to lookup in a reference table df_ref another value, like score info. the lookup is range-based, i.e. [0-10[, [10-20[ etc...and depends on the category (i.e each category has its own range and score info)
ex:
df_ex['category']=['A','B','A','B','B','A']
df_ex['score']=[1,45,65,7,34,76]
*********************************************
df_ref['category']=['A','A','A','A','B','B','B']
df_ref['low_bound']= [0,25,50,75,0,33,66] # >=
df_ref['up_bound']= [25,50,75,100,33,66,100] # <
df_ref['score_info']= ['low','medium','high','very high','low','medium','high']
and a magic_lookup on df_ex would then return ['low','medium','high','low', 'medium','very high'].
I see the nice answer from Joe here Best way to join / merge by range in pandas using Numpy broadcast.
I wonder how to generalize this when having the category criteria in addition.
Related
Lets say i have Dataframe, which has 200 values, prices for products. I want to run some operation on this dataframe, like calculate average price for last 10 prices.
The way i understand it, right now pandas will go through every single row and calculate average for each row. Ie first 9 rows will be Nan, then from 10-200, it would calculate average for each row.
My issue is that i need to do a lot of these calculations and performance is an issue. For that reason, i would want to run the average only on say on last 10 values (dont need more) from all values, while i want to keep those values in the dataframe. Ie i dont want to get rid of those values or create new Dataframe.
I just essentially want to do calculation on less data, so it is faster.
Is something like that possible? Hopefully the question is clear.
Building off Chicodelarose's answer, you can achieve this in a more "pandas-like" syntax.
Defining your df as follows, we get 200 prices up to within [0, 1000).
df = pd.DataFrame((np.random.rand(200) * 1000.).round(decimals=2), columns=["price"])
The bit you're looking for, though, would the following:
def add10(n: float) -> float:
"""An exceptionally simple function to demonstrate you can set
values, too.
"""
return n + 10
df["price"].iloc[-12:] = df["price"].iloc[-12:].apply(add10)
Of course, you can also use these selections to return something else without setting values, too.
>>> df["price"].iloc[-12:].mean().round(decimals=2)
309.63 # this will, of course, be different as we're using random numbers
The primary justification for this approach lies in the use of pandas tooling. Say you want to operate over a subset of your data with multiple columns, you simply need to adjust your .apply(...) to contain an axis parameter, as follows: .apply(fn, axis=1).
This becomes much more readable the longer you spend in pandas. 🙂
Given a dataframe like the following:
Price
0 197.45
1 59.30
2 131.63
3 127.22
4 35.22
.. ...
195 73.05
196 47.73
197 107.58
198 162.31
199 195.02
[200 rows x 1 columns]
Call the following to obtain the mean over the last n rows of the dataframe:
def mean_over_n_last_rows(df, n, colname):
return df.iloc[-n:][colname].mean().round(decimals=2)
print(mean_over_n_last_rows(df, 2, "Price"))
Output:
178.67
I have two columns: age and n_age. n_age is a number; age is a category of decade. For example, age might be 56, and n_age would be 50s.
I've filled the NAs of the n_age column with random sampling. I'd like to now use the most efficient method to fill the age column. I've used an apply with a lambda, like so:
df['age'].fillna(df['n_age'].apply(lambda x: str(x//10)+'0s'))
Thankfully my data set is relatively small, so this didn't take a long time. I imagine that there's a more efficient way of doing this--what is the better way?
No need apply
(df['age'].fillna(df['n_age'])//10).mul(10).astype(str)+'s'
I have a companies dataset with 35 columns. The companies can belong to one of 8 different groups. How do I for each group create a new dataframe which subtract the mean of the column for that group away from the original value?
Here is an example of part of the dataset.
So for example for row 1 I want to subtract the mean of BANK_AND_DEP for Consumer Markets away from the value of 7204.400207. I need to do this for each column.
I assume this is some kind of combination of a transform and a lambda - but cannot hit the syntax.
Although it might seem counter-intuitive for this to involve a loop at all, looping through the columns themselves allows you to do this as a vectorized operation, which will be quicker than .apply(). For what to subtract by, you'll combine .groupby() and .transform() to get the value you need to subtract from a column. Then, just subtract it.
for column in df.columns:
df['new_'+column] = df[column]-df.groupby('Cluster')['column'].transform('mean')
Help with homework problem: "Let us define the "data science experience" of a given person as the person's largest score among Regression, Classification, and Clustering. Compute the average data science experience among all MSIS students."
Beginner to coding. I am trying to figure out how to check amongst columns and compare those columns to each other for the largest value. And then take the average of those found values.
I greatly appreciate your help in advance!
Picture of the sample data set: 1: https://i.stack.imgur.com/9OSjz.png
Provided Code:
import pandas as pd
df = pd.read_csv("cleaned_survey.csv", index_col=0)
df.drop(['ProgSkills','Languages','Expert'],axis=1,inplace=True)
Sample Data:
What I have tried so far:
df[data_science_experience]=df[["Regression","Classification","Clustering"]].values.max()
df['z']=df[['Regression','Classification','Clustering']].apply(np.max,axis=1)
df[data_science_experience]=df[["Regression","Classification","Clustering"]].apply(np.max,axis=1)
If you want to get the highest score of column 'hw1' you can get it with:
pd['hw1'].max(). this gives you a series of all the values in that column and max returns the maximum. for average use mean:
pd['hw1'].mean()
if you want to find the maximum of multiple columns, you can use:
maximum_list = list()
for col in pd.columns:
maximum_list.append(pd[col].max)
max = maximum_list.max()
avg = maximum_list.mean()
hope this helps.
First, you want to get only the rows with MSIS in the Program column. That can be done in the following way:
df[df['Program'] == 'MSIS']
Next, you want to get only the Regression, Classification and Clustering columns. The previous query filtered only rows; we can add to that, like this:
df.loc[df['Program'] == 'MSIS', ['Regression', 'Classification', 'Clustering']]
Now, for each row remaining, we want to take the maximum. That can be done by appending .max(axis=1) to the previous line (axis=1 because we want the maximum of each row, not each column).
At this point, we should have a DataFrame where each row represents the highest score of the three categories for each student. Now, all that's left to do is take the mean, which can be done with .mean(). The full code should therefore look like this:
df.loc[df['Program'] == 'MSIS', ['Regression', 'Classification', 'Clustering']].max(axis=1).mean()
I'm trying to perform calculations based on the entries in a pandas dataframe. The dataframe looks something like this:
and it contains 1466 rows. I'll have to run similar calculations on other dfs with more rows later.
What I'm trying to do, is calculate something like mag='(U-V)/('R-I)' (but ignoring any values that are -999), put that in a new column, and then z_pred=10**((mag-c)m) in a new column (mag, c and m are just hard-coded variables). I have other columns I need to add too, but I figure that'll just be an extension of the same method.
I started out by trying
for i in range(1):
current = qso[:]
mag = (U-V)/(R-I)
name = current['NED']
z_pred = 10**((mag - c)/m)
z_meas = current['z']
but I got either a Series for z, which I couldn't operate on, or various type errors when I tried to print the values or write them to a file.
I found this question which gave me a start, but I can't see how to apply it to multiple calculations, as in my situation.
How can I achieve this?
Conditionally adding calculated columns row wise are usually performed with numpy's np.where;
df['mag'] = np.where(~df[['U', 'V', 'R', 'I']].eq(-999).any(1), (df.U - df.V) / (df.R - df.I), -999)
Note; assuming here that when any of the columns contain '-999' it will not be calculated and a '-999' is returned.