Skip to content

Instantly share code, notes, and snippets.

View meghankane's full-sized avatar

Meghan Kane meghankane

View GitHub Profile
View rf-distance-visualizations.R
if(!require("curl")) install.packages("curl")
if(!require("remotes")) install.packages("remotes")
# Based off of examples from RF documetation in TreeDist package
meghankane /
Last active March 5, 2023 18:47
Calculate Robinson-Foulds distance between 2 phylogenetic trees (given their newick strings as input)
from ete3 import Tree
import sys
# To Run: python3 path/to/our_method_results_file path/to/benchmark_results_file
# our method results file is a .log file & benchmark results file is a bestTree.tre
def parse_newick_string(filename_our_method, filename_benchmark):
newick_our_method = ""
newick_raxml_benchmark = ""
with open(filename_our_method, 'r') as file_our_method:
View gist:2cd53a832535d42af85f5100c2972dd9
### Keybase proof
I hereby claim:
* I am meghaphone on github.
* I am mkane ( on keybase.
* I have a public key ASAOfbSvzoop12nsm61GArVYJNkQ1a-hVFW_452TAKB9qAo
To claim this, I am signing this object:
View The State of ML for iOS... On the Advent of WWDC 2018 🕯🤓
The State of ML for iOS... On the Advent of WWDC 2018 🕯🤓
Machine learning should be accessible to all Apple developer community members who are curious and committed to learning. Let’s get everyone on the same page about the state of ML for iOS before WWDC, so we can hit the ground running next week!*
Core ML, coremltools, TuriCreate, Swift for TensorFlow, Metal Performance Shaders CNN, etc 🤯. There have been many exciting tools made available for developers in the recent past, so let’s review them:
- Demystify & disambiguate: What exactly are each of these?
- Workflow practicality: For what domain problems is it practical to use them?
- Implications: What engineering constraints are linked to using them?
- Big picture: How do the tools fit together in the larger picture of ML for on device prediction?
meghankane /
Last active May 9, 2018 13:06
import tfcoreml as tf_converter
tf_converter.convert(tf_model_path = 'retrained_graph.pb',
mlmodel_path = 'FlowerClassifierRetrained.mlmodel',
output_feature_names = ['final_training_ops/Wx_plus_b/add:0'],
image_input_names = 'input:0',
class_labels = 'retrained_labels.txt',
red_bias = -1,
green_bias = -1,
blue_bias = -1,
meghankane / FileProviderEnumerator.swift
Last active September 29, 2017 13:01
FileProviderEnumerator - fetched items in enumerateItems not displaying in Document Browser
View FileProviderEnumerator.swift
As a first step for the File Provider Extension, I'm trying to display the files fetched from in the Document Browser.
From what I understand from the WWDC 2017 File Provider Enhancements video, these files are fetched in the `enumerateItems(for observer...` function.
And, if the enumerator is the root container, they are fetched from the server ( in my case).
When debugging, the files are fetched successfully and stored in `fileProviderItems`.
I thought the remaining steps were to call `observer.didEnumerate(fileProviderItems)` and `observer.finishEnumerating(upTo: nil)`.
However, no luck with them displaying in the document browser.
Is there something obvious that I am missing in enumerateItems? Thanks again for your help!
meghankane /
Last active June 6, 2017 14:36
Bringing Machine Learning to your iOS Apps 🤖📲

Machine learning is a vast topic, so it can seem daunting to get started. However, as iOS app developers, we do not need to be experts to utilize the ML tools available to enhance our apps. Using machine learning in our apps can help us better understand our users so that we can build a better product for them 🔨😎.

This talk will cover:

  1. Brief overview of foundational ML concepts (neural networks, training, inference)
  2. Examples of how machine learning could help enhance your app and when it is practical
  3. What’s available from Apple - what’s new from WWDC 2017 (CoreML!!), practical overview of Metal, and how to get started using these APIs quickly
  4. Workflow and architecture for using ML
View Bug rating system (severity & occurrence)
Label each bug with the following two classifications
1. Severity: given the issue occurred, what is its impact on the user or the business
sev-1 maximal: Blocks application usage, a failure would permanently damage the reputation of the product and cause users to stop using product.
sev-2 considerable: A failure would block usage scenarios.
sev-3 some: A failure that might annoy the user. Retry or recovery mechanisms are straightforward.
sev-4 minimal: A failure the user might not even notice.
2. Occurrence: how likely is it for the issue to happen

Keybase proof

I hereby claim:

  • I am meghaphone on github.
  • I am meghafon ( on keybase.
  • I have a public key whose fingerprint is 0985 72C4 9466 75FC D8C4 204A 5E1A FCC5 10D0 6CFA

To claim this, I am signing this object: