Dear all,
I'm willing to perform a Linear Discriminant Analysis in Python using my own dataset. However the tutorials I find online only give examples of python's dataset. Can anyone please help me?
Kindly,
Cesar
Hello César António Mubango Hoguane ,
Have a look at this walk-through (using Fisher's iris data as an example): https://www.statology.org/linear-discriminant-analysis-in-python/
Obviously, you'll need to construct your own data frame that holds the data points you wish to evaluate, and that the appropriate type (e.g., numeric for IVs) is used. Have a look at this explanation: https://www.w3schools.com/datascience/ds_python_dataframe.asp
Good luck with your work.
Are you able to manipulate your dataset in such a way that it matches the structure presented in the tutorials? Whats the issue specifically?
Mr. David Morse I appreciate your answer. I'll detailly to these websites in order to perform this analysis.
Mr. Christian Schmidt I'm not sure whether I'm able to able to manipulate my dataset in such way. I want to make this analysis to discover whether the sediments of an Estuary are from the river or from the Ocean. I read a paper in which they make this analysis to infer about this.
César António Mubango Hoguane Which is the data imput? You can compute it using nominal sieve openings and cumulative mass of material by using Linear Discriminant adopted from Sahu (1964). It is s a powerful method to determine sedimentation process.
Sahu, B. K. (1964) Depositional mechanisms from the size analysis of clastic sediments, Journal of Sedimentary Petrology, Vol. 34, No. 1, pp. 73-83.
med = the input is the given grain 'media'
des = the input is the given grain 'sorting'
ses = the input is the given grain 'skewness'
cur = the input is the given grain 'kurtosis'
Yab = (-3.5688 * med) + (3.7016 * des**2) - (2.0766 * ses) + (3.1135 * cur)
Ysa = (15.6534 * med) + (65.7091 * des**2) + (18.1071 * ses) + (18.5043 * cur)
Yds = (0.2852 * med) - (8.7604 * des**2) - (4.8932 * ses) + (0.0482 * cur)
Yft = (0.7215 * med) + (0.403 * des**2) + (6.7322 * ses) + (5.2927 * cur)
if Yab -2.7411:
Yab_eti = print('Yab_Beach environment')
if Ysa 63.3650:
Ysa_eti = print('Ysa_Shallow agitated marine environment')
if Yds -7.419:
Yds_eti = print('Yds_Shallow marine deposit')
if Yft 10:
Yft_eti = print('Yft_Turbidity environment')
The link bellow leads to a code that will provide the input variables for the sahu method: mean, sorting (standard deviation), skewness and kurtosis using the logarithmic method described by Folk and Ward (1957).
https://github.com/pjparis/pySSAS-python-Sediment-Size-Analysis-by-Sieve
Folk, R.L., and Ward, W.C. (1957) Brazos River bar, a study in the significance of grainsize parameters, Journal of Sedimentary Petrology, Vol. 27, pp. 3- 26
for each sediment data collection point you can discriminate between different mechanisms and different deposition environments based on this method. But you can better automate the process in the language you want to generate the respective graphics.
Yours sincerely
- Badges
- Science topic