Find 4 Values In Window Size 6 That Fit Criteria -> Add To List Until 3 Do Not Fit Criteria And Do Not Include Last 3 -> Repeat Where It Left Off . Chou Fasman . Python

if you have a list of values:

values=['130','90','150','123','133','120','160','45','67','55','34','130','120','180','130','10']

and wanted to scan through with a window size of 6 and if 4 out of the 6 were >= 100 then keep scanning until there were 3 in a row that were < 100 and then not include those in the list

so for example with an empty list called results:

results=[]

i would like to append those values that satisfied the criteria into the empty list to get

results=[('130','90','150','123','133','120','160'),('55','34','130','120','180','130','10')]

i know i have convert all the strings into integers with int() but that's not the part that i'm having trouble with. i'm having trouble finding the 4 out of the window size 6 that are >= 100, adding that to a list, AND THEN going from where i left off in the window

In the Chou Fasman Algorithm each window size is 6 and if 4 are greater than 100 then all 6 are included and its extended until 4 consecutive values (i'm going to do 3 instead) are less than 100 (those 4 (or 3) are not included) and then the window starts again from that spot making a new list.

so first window would be:

['130','90','150','123','133,'120'] #and more than 4 are greater than 100 so that starting point is stored and the next window is checked
['90','150','123','133','120','160'] #again there are 4 greater than 100 so the next window is checked
['150','123','133','120','160','45'] #again
['123','133','120','160','45','67'] #again
['133','120','160','45','67','55'] #stop and assign values '130' to '160' into a list and then start the new window from where it left off

Results=[('130','90','150','123','133','120','160')]

['120','160','45','67','55','34'] # skips
['160','45','67','55','34','130'] # skips
['45','67','55','34','130','120'] # skips
['67','55','34','130','120','180'] # skips
['55','34','130','120','180','130'] # new list in Results starts with '55'
['34','130','120','180','130','10'] # sequence ends and this window still fits criteria so include these into the list so the results would now be

Results=[('130','90','150','123','133','120','160'),('55','34','130','120','180','130','10')]

I'd really like stay clear of yields and generators if possible

python windows function

• 1,818 views • link

updated 14.7 years ago by xenophiliuslovegood •

• written 14.7 years ago by hicsuntdrac0nis •

For questions related to programming you should consider them posting at stackoverflow.

• 0 views • link

written 14.7 years ago by Dataminer

Can you give the question some biological relevance? I suspect there is some application, but you can see that without mentioning it our edification suffers, and the moderators get pissed off (and close your question).

• 0 views • link

written 14.7 years ago by seidel

Albeit not very prominently, the question does mention the Chou-Fasman method, so I don't see it as being particularly off topic.

• 0 views • link

written 14.7 years ago by xenophiliuslovegood •

@xenophiliuslovegood: In that case we should open a new forum to discuss programming issues related to bioinformatics. ;)

• 0 views • link

written 14.7 years ago by Dataminer

4 answers

This will do the work

start=False
results=[]
allresults=[]
for x in range(len(values)-6):
    if start or sum(int(x)>=100 for x in values[x:x+6])>4: # begin appending if window condition satified
        start=True
    if start:
        results.append(values[x])
        if sum(int(x)<100 for x in values[x+1:x+5])>=4: # break loop if next 4 are less than 100
            allresults.append(results)
            results=[]
            start=False

if start:
    allresults.append(values[x:])

print (str(allresults))

• 0 views • link

written 14.7 years ago by Farhat

ahh almost i change value to values=['130','90','150','123','133','120','160','180','45','67','55','34','130','120','180','130','10','30'] and it should spit out results=[('130','90','150','123','133','120','160','180'),('55','34','130','120','180','130','10','30')]

because in the Chou Fasman Algorithm each window size is 6 and if 4 are greater than 100 then all 6 are included and its extended until 4 consecutive values are less than 100 and then the window starts again making a new list. thanks tho this definitely set me in the right direction i'm going to work with this . been over 6hrs

• 0 views • link

written 14.7 years ago by hicsuntdrac0nis •

i changed the question to make it easier and more specific

• 0 views • link

written 14.7 years ago by hicsuntdrac0nis •

Instead of the break, append the results to another list and clear the results list. This should get you all the results instead of just the first one.

• 0 views • link

written 14.7 years ago by Farhat

Changed the program to incorporate that.

• 0 views • link

written 14.7 years ago by Farhat

Changed the program to incorporate some of that. You could modify some of the values to get the specific results you want.

• 0 views • link

written 14.7 years ago by Farhat

So basically you just want to remove any series of numbers that are all less than 100 and appear more than 4 times in a row?

Something like this should work:

values=['130','90','150','123','133','120','160','180','45','67','55','34','130','120']

less = []
result = []
for i in values:
    if int(i) < 100:
        less.append(i)
    else:
        if len(less) < 4:
            for j in less:
                result.append(j)
            less = []
        else:
            less = []
        result.append(i)

print result

• 1 views • link

written 14.7 years ago by Damian Kao

almost . i want it to end and create a list out of the values that it found . so if it had the values values=['130','90','150','123','133','120','160','180','45','67','55','34','130','120','180','130','10','30'] it should make the results=[('130','90','150','123','133','120','160','180'),('55','34','130','120','180','130','10','30')] because in the Chou Fasman Algorithm each window size is 6 and if 4 are greater than 100 then all 6 are included and its extended until 4 consecutive values are less than 100 and then the window starts again making a new list. thanks tho this definitely set m

• 0 views • link

written 14.7 years ago by hicsuntdrac0nis •

i changed the question to make it easier and more specific

• 0 views • link

written 14.7 years ago by hicsuntdrac0nis •

It's not clear to my why your second tuple in all-results include that final '10' as that seems to go against your description. I have implemented a solution using generators ('cause that's the easiest way), but the interface returns a list. The parameters are adjustable, so you can change the various cutoffs.

def _chou_fasman(values, window_size, cutoff, n_above, n_below):
    i = -1
    while i < len(values) - window_size:
        i += 1
        current = values[i: i + window_size]
        if sum(1 for c in current if c > cutoff) < n_above: continue
        j = i + window_size + n_below
        while sum(1 for v in values[i + window_size: j] if v < cutoff) < n_below and j < len(values):
            j += 1
        last_index = j - n_below # you may need to change this, it's not clear from you example.
        yield values[i:min(len(values), last_index)]
        i = last_index

def chou_fasman(values, window_size=6, cutoff=100, n_above=4, n_below=3):
    return list(_chou_fasman(values, window_size, cutoff, n_above, n_below))

if __name__ == "__main__":
    values = map(int, ['130','90','150','123','133','120','160','45','67',
                       '55','34','130','120','180','130','10'])
    print "\n".join(map(str, chou_fasman(values)))

where you'll call chou_fasman so as not to have to deal with generators.

• 1 views • link

written 14.7 years ago by brentp

there is a typo in his example solution. It should not be ('55','34','120','180','130','10') rather [55, 34, 130, 120, 180, 130, 10] (a 130 is missing from his list).

• 0 views • link

written 14.7 years ago by xenophiliuslovegood •

why is there a 10 on the end?

• 0 views • link

written 14.7 years ago by brentp

@ xenophiliuslovegood : i fixed the typo you were right @brentp : the 10 is on the end b/c the last window of size 6 includes the 10 and that window still has at least 4 values greater than 100 so the last window is accepted in the algorithm

• 0 views • link

written 14.7 years ago by hicsuntdrac0nis •

As far as I understand this does what you want, produces the correct answers when fed with all the examples you provided (except I think there's a typo in one of your example solutions), certainly does not use generators.

def build_choufasman_regions(l):
    le=len(l)
    tmp=[];res=[]
    pos=0
    while (pos<= (le-6) ):
        c=0;
        for i in range(pos,pos+6):
            if (l[i]>=100): c=c+1
        if c>=4:
            if ([]==tmp):
                tmp.extend(l[pos:pos+6])
            else:
                tmp.extend([l[pos+5]])
        else:
            if ([]==tmp): pass
            else:
                if (tmp[-1]<100 and tmp[-2]<100):
                    tmp.pop(-1);tmp.pop(-1);pos=pos+3
                res.append(tmp)
                tmp=[]
        #print pos,tmp
        pos=pos+1
    if ([]!=tmp):res.append(tmp) 
    return res

l=[130,90,150,123,133,120,160,45,67,55,34,130,120,180,130,10]   
print build_choufasman_regions(l)
l=[130,90,150,123,133,120,160,180,45,67,55,34,130,120,180,130,10,30]
print build_choufasman_regions(l)
l=[1,1,1,200,200,200,100,100,50,100]
print build_choufasman_regions(l)

• 0 views • link

written 14.7 years ago by xenophiliuslovegood •

I fixed the formatting of the code, indentation had come out wrong while copy/pasting.

• 0 views • link

written 14.7 years ago by xenophiliuslovegood •