Will display the image in your default browser.

Will calculate the required percentile range for each slice of data. :return: An array containing the calculated percentiles.

Will generate a graph object which can be used to verify the results. :return: The plotly graph object/

Will render the data into a figure. :return:

Will render the data into a figure. :return:

A simulation where the benchmark is consistently randomly increased. This will generate an ever changing benchmark that can help us f

:return:

Saving image using the orca engine the default kaleido engine was not working for me. :param image_format : The format of the

the Kullback–Leibler divergence is used to represent a measure of distance between a percentile range from A to B. More info

Will use the kolmogorov smirnov statistical test to calculate the distance between two ECDF distributions. The KS test measures how

Used to calculate the percentile over the given array. :return: the requested percentile as a float

Will calculate the distance score by looking at how many measurements exhausted its threshold. :return: The distance score

Computes the Wasserstein distance or Kantorovich–Rubinstein metric also known as the Earth mover's distance. This metric represents h

Will create a scenario based on the scenario information. :param percent_of_data_set: The amount in percentage of the data se

Will estimate the change value from A to B using the Kullback–Leibler divergence. This value can be considered as the absolu

Will generate a graph object which can be used to verify the results. :return: The plotly graph object.

Will generate a scoring matrix that can be used to make a score go up or down. It does this by using the dirichlet distribution.

Will go through each percentile measurement and calculate the percentage difference between them which it will push through its respe

Will give the letter grade to the D value score. Below a rough estimate how much regression in percentage each letter grade r

An heuristic that will estimate a rank of what the amount of change is between our distributions. This rank is based on the Japanese

Will iterate a value trough the negative change grid and will punish the score when a threshold is broken. :param value: The

Will iterate a value trough the positive change grid and will punish the score when a threshold is broken. :param value: The

Will compare the scenario using the distance test and returning the metrics. :param scenario: the simulated scenario containing all o

Will execute the T-test. - False: We can reject the null hypothesis the test is slower or faster than the baseline. - True: W

Will use a change matrix to score every Kullback–Leibler divergence statistics grading it from 0 to 1. :return: retu

An heuristic that will estimate a score between 0 - 100 using the Wasserstein distance and the kolmogorov smirnov distance. I

Will run a simulation. :param percent_of_data_set: The amount in percentage of the data set that needs to be changed.

Will check if both arrays equal zero or not/ :return: True: There is no change both sums are equal to each other or - False:

Will transfer the lines of the chunk into the json data set. Keep in mind that the json structure will be kept in the correct

Will calculate the eCDF to find the empirical distribution of our population. This function will then create a dataframe which will contain t

Will calculate the percentage change. :param old: The old measurement. :param new: the new measurement. :return: The percentage chang

Will read the csv file in chunks and convert it to json.

Will normalize a given raw distribution to its maximum. param data: A list of raw measurements that need to be normalized. param percenti

:param delta: :param population: :param percentage: :param positive: True when the change the delta needs to increas

Will construct the class and calculate all the required statistics. :param population_a: An list of floats of the A population (base

Will set up th class and find the max edge of the score from which the distance will be calculated by determining the length of the b

Will construct the class and calculate all the required statistics. After all the computation have been completed the following infor

Will load in the object and map the object arguments to attributes. These attributes ares used to identify the data. :param

When constructed will start the conversion process. :param path: The path to the file

:param percentage: :param delta: :param baseline_id: :param benchmark_id: :param positive: True when the cha

Will build the image. :param benchmark: :param baseline:

will build a scatter plot image :param scenario: :param rank: :param change:

Will construct the simulation object and provide a few attributes which can be changed. :param benchmark_id: The RID that ne

The average metric calculated over all of the measurements. :return: a float that represents the average of the sample

Will count how many measurements there are loaded into this object. In the performance engineering context this count

Will calculate the critical t-value. :return: The critical t-value

The ECDF (empirical cumulative distribution function) Will calculate the eCDF to find the empirical distribution of our population.

will convert the python dictionary to true json :return:

Computes the absolute distance between 2 distribution representing the maximum distance between sample A and B. More informat

Not used in any of the calculation in this heuristic but it might be helpful for engineers that also want to perform a official KS-te

Will find out the minimum outlier from all of the collected measurements (raw data). :return: A float that is the maximum num

Will calculate the median over the collected measurements. (raw data) :return: A float which is the standard deviation.

Will find out the maximum outlier from all of the collected measurements (raw data). :return: A float that is the maximum num

Will normalize a given raw distribution to its maximum. When normalizing it wil exclude all measurement below the 95th percentile.

Will calculate the ECDF (empirical cumulative distribution function) to find the empirical distribution of our population. It

Will calculate a pre-defined set of percentiles from the raw data. :return: Will return the percentile measurement as a tuple.

Complex classes can confuse QuickPotato and make it hard to find a starting point from where QuickPotato can start unwrapping

This property represent the collected data in its raw format. It is used by this object itself to calculate other metrics and informa

Will create gif. :return:

The results of the t-test. - False: We can reject the null hypothesis the test is slower or faster than the baseline.

Saving image using the orca engine the default kaleido engine was not working for me. :param image_format : The format of the

A score from 0 to 100 that represents the amount of change between two percentile distributions. :return: The classic distanc

Will structure the raw data object used in the tests.

Will display the image in your default browser.

Will calculate the standard deviation over the collected measurements. (raw data) :return: A float which is the standard devi

This total amount of all of the measurement combined. :return: returns an integer representing the total sum of all measureme

Will calculate the t-value. :return: The calculate t-value

Verifying if it is possible to find the correct rank and distance metrics for one of my dummy runs. Be aware that distance metrics ca

A simple code performance test which allows me to access complex code performance visualizations and test if my heuristic follows

Check if a boundary of a letter rank based on the pre-defined boundary matrix. By matching the letter rank to a number determining if a lette

Will check the boundary of a change value by validating if the value - breaches the max or min boundary. :param change: The change value

Gives the computed value of distance from sample a versus sample b. This value represent how much effort there is required to move th