spelling updates

README.md

@@ -23,8 +23,8 @@ alfie -h
 ```
 Or open python and run the following
 ```
-from alfie import seqio
-?seqio
+from alfie import seqio #import the seqio module
+?seqio					#check you can see the help menu
 ```
 
 To install the most recent (development) version of alfie, download and unzip this repository. From the terminal, enter the repository and then run the following command:

alfie/kmerseq.py

@@ -49,7 +49,7 @@ class KmerFeatures:
 	Examples
 	---------
 	#initiate a class instance
-	#by kmer counts are generated on initialization, by default for k = 4.
+	#by default kmer counts are generated on initialization for k = 4.
 	>>> ex_inst = KmerFeatures(name = 'ID1', sequence = 'AAATTTGGGATGGGCCCCACAC')
 
 	# numpy array with the kmer labels

example/.ipynb_checkpoints/custom_alfie_demo-checkpoint.ipynb

@@ -115,7 +115,7 @@
    "source": [
     "The classification process is completed above in just one line of code. In the background the `classify_records` function is taking the input sequences, generating a set of input features (k-mer frequencies) for each sequence, and passing the feature sets through a neural network to obtain a prediction. \n",
     "\n",
-    "The function yields two outputs, a list of sequence records and an array classifications. The sequence records are a list of dictionaries, like the inputs, but now with an additional 'kmer_data' entry which contains a class instance used to generate the input features for the neural network (more information on the `KmerFeatures` class is provided in the following section on training a classifier)."
+    "The function yields two outputs, a list of sequence records and an array of classifications. The sequence records are a list of dictionaries just like the inputs. Now there is an additional 'kmer_data' entry, which contains a data structure used to generate the input features for the neural network (more information on the `KmerFeatures` class is provided in part 2.)."
    ]
   },
   {
@@ -175,7 +175,7 @@
    "cell_type": "markdown",
    "metadata": {},
    "source": [
-    "For some purposes, the names corresponding to the encoded predictions may be preferable to the numeric encodings. To obtain these, use the `decode_predictions` to move from the array of numeric predictions to a list of names.  "
+    "For some purposes, the names corresponding to the encoded predictions may be preferable to the numeric encodings. To obtain these, use the `decode_predictions`function to move from the array of numeric predictions to a list of names.  "
    ]
   },
   {
@@ -205,7 +205,7 @@
    "cell_type": "markdown",
    "metadata": {},
    "source": [
-    "Working within python provides the freedom to manipulate the sequence records in various ways using this information. What you do with the classification information will depend on your research goal, you may wish to save a select category of sequences to a file, merge the classifications with the existing sequence ids, or carry the classifications through to additional analyses in python. Here we explore a few of these possibilities.\n",
+    "Working within python provides the freedom to manipulate the sequence records in various ways using this information. What you do with the classification information will depend on your research goal. You may wish to save a select category of sequences to a file, merge the classifications with the existing sequence ids, or carry the classifications through to additional analyses in python. Here we explore a few of these possibilities.\n",
     "\n",
     "\n",
     "**a.** Add the predictions into the sequence dictionaries prior to subsequent manipulation."
@@ -1180,7 +1180,7 @@
    "cell_type": "markdown",
    "metadata": {},
    "source": [
-    "Below the classifier is trained in an extremely simple fashion, the data is passed through the model in batches of 100 for 50 epochs. We could improve this model with cross validation and hyperparameter tuning (i.e. changing the dropout or the number of hidden layers). Additionally you could change the set of input features (size of k), or take lower level control and design a neural network yourself. These refinements and optimizations are outside of the scope of the alfie tutorial. Scikit learn and tensorflow have excellent functions and documentation if you wish to branch out beyond the `alfie_dnn_default` architecture."
+    "Below the classifier is trained in an extremely simple fashion, the data is passed through the model in batches of 100 for 50 epochs. We could improve this model with cross validation and hyperparameter tuning (i.e. changing the dropout or the number of hidden layers). Additionally you could change the set of input features (size of k), or take lower level control and design a neural network yourself. These refinements and optimizations are outside of the scope of the alfie tutorial, but will offer meaningful improvement in most situations. Also scikit learn and tensorflow have excellent functions and documentation if you wish to branch out beyond the `alfie_dnn_default` architecture."
    ]
   },
   {
@@ -1536,7 +1536,7 @@
    "cell_type": "markdown",
    "metadata": {},
    "source": [
-    "The custom models can be used from outside of python as well with the alfie command line function! Simply specify the custom model with the `-m` flag and the custom classes with the `-c` flag to conduct custom file-to-file classification."
+    "The custom models can be used from outside of python as well in conjunction with the alfie command line function! Simply specify the custom model with the `-m` flag and the custom classes with the `-c` flag to conduct custom file-to-file classification."
    ]
   },
   {

example/custom_alfie_demo.ipynb

@@ -115,7 +115,7 @@
    "source": [
     "The classification process is completed above in just one line of code. In the background the `classify_records` function is taking the input sequences, generating a set of input features (k-mer frequencies) for each sequence, and passing the feature sets through a neural network to obtain a prediction. \n",
     "\n",
-    "The function yields two outputs, a list of sequence records and an array classifications. The sequence records are a list of dictionaries, like the inputs, but now with an additional 'kmer_data' entry which contains a class instance used to generate the input features for the neural network (more information on the `KmerFeatures` class is provided in the following section on training a classifier)."
+    "The function yields two outputs, a list of sequence records and an array of classifications. The sequence records are a list of dictionaries just like the inputs. Now there is an additional 'kmer_data' entry, which contains a data structure used to generate the input features for the neural network (more information on the `KmerFeatures` class is provided in part 2.)."
    ]
   },
   {
@@ -175,7 +175,7 @@
    "cell_type": "markdown",
    "metadata": {},
    "source": [
-    "For some purposes, the names corresponding to the encoded predictions may be preferable to the numeric encodings. To obtain these, use the `decode_predictions` to move from the array of numeric predictions to a list of names.  "
+    "For some purposes, the names corresponding to the encoded predictions may be preferable to the numeric encodings. To obtain these, use the `decode_predictions`function to move from the array of numeric predictions to a list of names.  "
    ]
   },
   {
@@ -205,7 +205,7 @@
    "cell_type": "markdown",
    "metadata": {},
    "source": [
-    "Working within python provides the freedom to manipulate the sequence records in various ways using this information. What you do with the classification information will depend on your research goal, you may wish to save a select category of sequences to a file, merge the classifications with the existing sequence ids, or carry the classifications through to additional analyses in python. Here we explore a few of these possibilities.\n",
+    "Working within python provides the freedom to manipulate the sequence records in various ways using this information. What you do with the classification information will depend on your research goal. You may wish to save a select category of sequences to a file, merge the classifications with the existing sequence ids, or carry the classifications through to additional analyses in python. Here we explore a few of these possibilities.\n",
     "\n",
     "\n",
     "**a.** Add the predictions into the sequence dictionaries prior to subsequent manipulation."
@@ -1180,7 +1180,7 @@
    "cell_type": "markdown",
    "metadata": {},
    "source": [
-    "Below the classifier is trained in an extremely simple fashion, the data is passed through the model in batches of 100 for 50 epochs. We could improve this model with cross validation and hyperparameter tuning (i.e. changing the dropout or the number of hidden layers). Additionally you could change the set of input features (size of k), or take lower level control and design a neural network yourself. These refinements and optimizations are outside of the scope of the alfie tutorial. Scikit learn and tensorflow have excellent functions and documentation if you wish to branch out beyond the `alfie_dnn_default` architecture."
+    "Below the classifier is trained in an extremely simple fashion, the data is passed through the model in batches of 100 for 50 epochs. We could improve this model with cross validation and hyperparameter tuning (i.e. changing the dropout or the number of hidden layers). Additionally you could change the set of input features (size of k), or take lower level control and design a neural network yourself. These refinements and optimizations are outside of the scope of the alfie tutorial, but will offer meaningful improvement in most situations. Also scikit learn and tensorflow have excellent functions and documentation if you wish to branch out beyond the `alfie_dnn_default` architecture."
    ]
   },
   {
@@ -1536,7 +1536,7 @@
    "cell_type": "markdown",
    "metadata": {},
    "source": [
-    "The custom models can be used from outside of python as well with the alfie command line function! Simply specify the custom model with the `-m` flag and the custom classes with the `-c` flag to conduct custom file-to-file classification."
+    "The custom models can be used from outside of python as well in conjunction with the alfie command line function! Simply specify the custom model with the `-m` flag and the custom classes with the `-c` flag to conduct custom file-to-file classification."
    ]
   },
   {