Create a new GPFeatureLayer from data frame

I am a complete newbie to ArcGIS/ArcPy. I am creating a new Python Toolbox, that reads in a GPFeatureLayer, and also should return one. I read in the values of a GPFeatureLayer row by row. Then I perform some calculations on them, to obtain indices of rows that are important for me. This results in a data frame.

Now I want to create a new GPFeatureLayer that only contains the rows that are specified by my indices. I thought that there may be a function to create a new GPFeatureLayer from a df, but is there any other way how to get to the goal?

This is my code:

import arcpy
from arcgis.gis import GIS
from arcgis import features
from arcgis.features import FeatureLayerCollection
import arcgis
import pandas as pd
import matplotlib.pyplot as plt
import numpy as np
from sklearn.linear_model import LinearRegression
from sklearn.preprocessing import PolynomialFeatures
from sklearn.metrics import r2_score


class Toolbox(object):
    def __init__(self):
        """Define the toolbox (the name of the toolbox is the name of the
        .pyt file)."""
        self.label = "Toolbox"
        self.alias = "toolbox"

        # List of tool classes associated with this toolbox
        self.tools = [Tool]


class Tool(object):
    def __init__(self):
        """Define the tool (tool name is the name of the class)."""
        self.label = "ComparingSegments"
        self.description = ""
        self.canRunInBackground = False

    def getParameterInfo(self):
        """Define parameter definitions"""
        p0 = arcpy.Parameter(displayName="Input Feature Layer",
                             name="input_feature_layer",
                             datatype="GPFeatureLayer",
                             parameterType="Required",
                             direction="Input")

        p1 = arcpy.Parameter(displayName="Output Feature Layer",
                             name="output_feature_layer",
                             datatype="DEFeatureClass",
                             parameterType="Required",
                             direction="Output")

        params = [p0, p1]

        return params

    def isLicensed(self):
        """Set whether tool is licensed to execute."""
        return True

    def updateParameters(self, parameters):
        """Modify the values and properties of parameters before internal
        validation is performed.  This method is called whenever a parameter
        has been changed."""

        return

    def updateMessages(self, parameters):
        """Modify the messages created by internal validation for each tool
        parameter.  This method is called after internal validation."""
        return


    def execute(self, parameters, messages):
        """The source code of the tool."""

        input_feature_layer = parameters[0].valueAsText

        fields = ["OBJECTID", "Shape", "ORIG_FID", "ORIG_SEQ", "Shape_Length", "Z_Mean"]
        z_means = []
        shape = []
        orig_fid = []
        orig_seq = []
        shape_length = []
        object_id = []

        with arcpy.da.SearchCursor(input_feature_layer, fields) as cursor:
            for row in cursor:
                z_means.append(float(row[5]))
                shape.append(int(row[4]))
                orig_fid.append(int(row[3]))
                orig_seq.append(int(row[2]))
                shape_length.append(int(row[1]))
                object_id.append(int(row[0]))

        startIndex, endIndex = someFunction(z_means, object_id)

        data = {"OBJECTID": object_id[startIndex:endIndex],
                "Shape": shape[startIndex:endIndex],
                "ORIG_FID": orig_fid[startIndex:endIndex],
                "ORIG_SEQ": orig_seq[startIndex:endIndex],
                "Shape_Length": shape_length[startIndex:endIndex],
                "Z_Mean": z_means[startIndex:endIndex]}

        df = pd.DataFrame(data)

        # HOW TO CREATE AND RETURN NEW GPFeatureLayer FROM DF???

        return

    def postExecute(self, parameters):
        """This method takes place after outputs are processed and
        added to the display."""
        return result