DIG DEEPER IN MINUTES WITH THE 5 STEP API

The Best Estimate API can provides output for variability exercises with little work on your side. This implies that you can integrate reliability analyses directly in your workflow. In short:

  • Easily accessible, inexpensive, fast and simple to use
  • No vendor lock-in, your data & results remain yours
  • Easy to integrate in your own project or company due to API structure
  • Quality control in your hands due to reporting of intermediate output
Want to learn more or simplify and optimize your designs by means of clustering? Subscribe to the newsletter and use the API for free.


Disclaimer. Privacy. Unsubscribe. Reach out.

COMPLEX ANALYSES IN 5 STEPS

Automation of the hard work leaves you with the most valuable part of the work: interprete and use the results. Analyses which were out of reach are completed in minutes with the API-based interface.

  1. Define analyses you require to do based on the examples provided
  2. Prepare input file with data required as provided on the public repository
  3. Log in on the API server by providing your account details
  4. Get a coffee, but be quick, your analyses may finish in seconds!
  5. Save output report and logfile or continue your calculation with the results
import requests, json, base64

# 1. Define API location
url = https://bestestimate.nl/

# 2. Login to the web-API
login = requests.post(url+"login",json={"key":"file_your_api_key_here"})

# 3. Perform analysis and retrieve results
analysis = requests.post(url+"bootstrap",json=data,headers={"Authorization":str(login.json()['access_token'])})

# 4. Save output plot
create_image("report",analysis.json()["report_1_png"],analysis.json()["report_1_pdf"])

# 5. Save logfile
create_logfile("logfile",analysis.json()["log_string"])

INTERMEDIATE OUTPUT FOR QUALITY CONTROL

Every operation performed by the API is logged in a logfile. This has the benefit that you can trace back all individual steps executed if you want to go into detail. Apart from the logfile a seperate output file containing a library of all results is supplied. This allows integration in your workflow and avoids losing time with copying and pasting data. Below box shows the output for a full distribution fitting operation. Check the public repository for more examples which you can try yourself.

-------------------------------------------------------------
BEST ESTIMATE - ANALYSIS LOG
-------------------------------------------------------------

Analysis performed: Distribution fitting
Date: 2020-11-17
Start time: 15:56:03

-------------------------------------------------------------
DATA VALIDATION
-------------------------------------------------------------

15:56:03 : Data validation started
15:56:03 : Data input correct
15:56:03 : References from user added to the output sheet
15:56:03 : Bootstrap total equal to provided value of 10000
15:56:03 : Bootstrap size equal to provided value of 8
15:56:03 : Confidence interval as provided used: 95%
15:56:03 : Validation is completed, input is correct so analysis can be started

-------------------------------------------------------------
BASIC DATA ANALYSIS
-------------------------------------------------------------

15:56:04 : Size dataset: 30
15:56:04 : Mean value dataset: 34.53
15:56:04 : Minimum value dataset: 27.1
15:56:04 : Maximum value dataset: 40.0
15:56:04 : Lower bound: 28.62
15:56:04 : Upper bound: 39.42

-------------------------------------------------------------
BOOTSTRAP ANALYSIS
-------------------------------------------------------------

15:56:04 : Bootstrap analysis started
15:56:04 : Bootstrap mean: 34.64
15:56:04 : Bootstrap lower bound: 32.35
15:56:04 : Bootstrap upper bound: 36.78
15:56:04 : Histogram created with 5 bins

-------------------------------------------------------------
QUANTILE QUANTILE ANALYSIS
-------------------------------------------------------------

15:56:04 : Quantile-quantile analysis started
15:56:04 : Least square errors calculated...
15:56:04 : Normal Distribution : 15.86
15:56:04 : Uniform Distribution : 16.52
15:56:04 : Exponential Distribution : 37.67
15:56:04 : Lognormal Distribution : 42.94
15:56:04 : Distribution best fitting : Normal Distribution

-------------------------------------------------------------
DISTRIBUTION FIT
-------------------------------------------------------------

15:56:04 : Distribution fitting of Normal Distribution started
15:56:04 : Parameter 1 : 34.53
15:56:04 : Parameter 2 : 3.23
15:56:04 : Confidence interval lower bound: 28.19
15:56:04 : Confidence interval upper bound: 40.86

-------------------------------------------------------------
ANALYSIS COMPLETED
-------------------------------------------------------------

Calculation log pushed in bytes file to user
Output image(s) pushed in bytes file to user

End time: 15:56:08